New Control Strategy Boosts Efficiency and Stability in DC Power Grids
The world’s power systems are undergoing a transformation. As renewable energy sources like solar and wind become more prevalent, and electric vehicles (EVs) reshape transportation and energy demand, traditional alternating current (AC) power grids face mounting challenges. Voltage instability, inefficiencies in power conversion, and difficulties in managing fluctuating loads are becoming increasingly common. In response, researchers are turning to direct current (DC) distribution systems, which offer a more efficient and stable alternative for modern energy needs. A recent breakthrough from a team at Tianjin University presents a novel control strategy that could significantly enhance the performance of medium-voltage DC (MVDC) grids, making them more economical and resilient, especially under extreme conditions like sudden power surges.
This new development, published in the prestigious Proceedings of the CSEE, comes at a critical time. The global push for decarbonization and the electrification of everything from cars to homes is placing unprecedented stress on existing infrastructure. Conventional power grids, designed for a different era, are struggling to keep up. The integration of distributed energy resources—such as rooftop solar panels and home battery systems—introduces a level of variability that traditional control methods were not built to handle. When a large data center suddenly increases its power draw, or when a passing cloud causes a rapid drop in solar generation, the resulting power imbalance can cause voltage fluctuations that threaten the stability of the entire system. These events, known as “special operating conditions,” can lead to inefficiencies, increased operational costs, and in the worst cases, equipment damage or power outages.
The research team, led by Professor Xiao Qian and doctoral candidate Lu Wenbiao from the Key Laboratory of Smart Grid of the Ministry of Education at Tianjin University, has developed a dual-time-scale hierarchical regulation and control method designed to solve these exact problems. Their approach is not just a minor tweak to existing technology; it represents a fundamental rethinking of how MVDC systems are managed, combining advanced scheduling with a smarter, more responsive control mechanism at the hardware level.
At the heart of their innovation is an improved version of what is known as “droop control.” In a DC grid, maintaining a stable bus voltage is paramount. If the voltage drops too low or rises too high, connected devices can malfunction. Droop control is a widely used technique that allows multiple power converters to share the load in a decentralized way. It works on a simple principle: as the power demand increases, the voltage is allowed to decrease slightly, and the converters automatically respond by increasing their output to compensate. This creates a self-regulating system without the need for a central command for every minor adjustment.
However, the researchers identified a critical flaw in conventional droop control. Under normal, small fluctuations in power, it works well. But when faced with a large, sudden change—like a data center’s load doubling in seconds—the traditional method fails. The voltage can drop so far that it violates safety limits before the system can respond, leading to instability. “The conventional droop curve is linear,” explained one of the team’s engineers in a recent interview. “It treats a 1% power increase the same way as a 50% increase. That’s like using the same brake pedal pressure to stop a car going 10 miles per hour as you would for one going 100. It’s simply not effective for extreme events.”
To overcome this, the Tianjin team introduced a revolutionary modification: they replaced the linear droop curve with one based on the mathematical tangent function, or “tan-function.” This might sound like a small change, but its implications are profound. The tan-function curve is flat in the middle, behaving almost identically to the traditional droop control during small, everyday fluctuations. This ensures high efficiency and optimal economic operation under normal conditions. However, when a large disturbance causes the voltage to deviate significantly from its target, the tan-function curve becomes extremely steep. In practical terms, this means that the power converter reacts with a massive, almost instantaneous increase in output power. This rapid response acts like a powerful safety clamp, preventing the voltage from falling below the critical threshold and keeping the system stable.
“This is the key innovation,” said Lu Wenbiao, the paper’s lead author. “Our improved droop control creates a seamless, automatic switch between two modes. Under normal conditions, it’s a droop controller, maximizing economic efficiency. When a crisis hits, it instantly becomes a voltage stabilizer, prioritizing system safety. It’s like having a car that drives smoothly and efficiently on the highway but can deploy its full braking power the moment it detects a sudden obstacle.”
The researchers didn’t stop at the hardware-level control. They embedded this new “tan-droop” control within a sophisticated two-layer management framework. The first layer operates on a long time scale, typically over an hour or more. In this layer, a central system scheduler uses forecasts of renewable energy generation, electricity prices, and load demands to create an optimal operating plan. The goal is to minimize the total operating cost. This involves decisions like when to charge a large battery storage system during cheap, off-peak hours and when to discharge it during expensive, high-demand periods—a strategy known as “low storage, high generation.” It also involves managing the charging schedules of EV charging stations, ensuring they meet contractual obligations while avoiding high-cost electricity.
The second layer operates on a much shorter time scale, reacting to events every few minutes or even seconds. This is where the improved droop control shines. While the long-term plan sets the target operating points, the short-term layer uses the tan-droop mechanism to handle the real-world chaos that forecasts can’t perfectly predict. If the wind suddenly dies down, or if a factory unexpectedly turns on a large machine, the tan-droop controllers on the power converters react instantly to maintain voltage stability, all while the system continues to follow the overall economic plan as closely as possible.
To test their theory, the team built a detailed computer model of a ±10 kV three-terminal ring-shaped MVDC system. This model included a complex network of components: three hybrid modular multilevel converters (MMCs) connecting to the AC grid, distributed photovoltaic (PV) and wind turbine (WT) generation, a large-scale energy storage system (ESS) with a 28 MW·h capacity, and multiple EV charging stations. The simulated grid also served a variety of loads, including an industrial facility, a residential area, a sensitive data center, and a “new-type industrial park” with its own local renewable generation.
The results of their simulation were compelling. The researchers compared five different operational scenarios. In one, the system used traditional, non-optimized droop control. In another, it used a fixed-power control strategy, which is simpler but less flexible. The baseline scenario used optimized droop control, which is considered state-of-the-art. The final scenario implemented their new dual-time-scale method with the tan-droop control.
The economic benefits were clear. When compared to a scenario where the energy storage system was not used optimally, the new method reduced the total operating cost by approximately 11%. This is a significant saving for a power system, translating to millions of dollars in real-world applications over time. The optimized use of the battery storage system was a major factor. The scheduler successfully charged the batteries during two low-price periods—early morning and mid-afternoon—and discharged them during two high-price periods, effectively “arbitraging” the electricity market to lower costs.
Even more impressive were the results regarding system stability. In a simulated extreme event, the load of the data center was increased by 100%—a massive, near-instantaneous power demand. In the scenario using traditional droop control, the DC bus voltage at the data center node plummeted, falling below the allowable safety limit of 97% of the rated voltage (19.40 kV). This voltage violation could have led to equipment shutdowns or damage in a real system. In stark contrast, when the new tan-droop control was activated, the voltage remained firmly within the safe operating range. The system responded so quickly and effectively that the voltage dip was minimal and never breached the critical threshold.
“The voltage stability under such a severe load surge is the most convincing evidence of our method’s superiority,” said Professor Xiao Qian. “It proves that our control strategy doesn’t just save money under normal conditions; it fundamentally improves the grid’s resilience and reliability when it matters most.”
The implications of this research extend far beyond the laboratory. As cities look to build more robust and sustainable energy infrastructure, MVDC grids are a promising solution. They are particularly well-suited for dense urban environments, industrial parks, and data center campuses, where high power demands and a high concentration of electronic loads make DC power a natural fit. The ability to seamlessly integrate solar panels, wind turbines, battery storage, and EV charging stations into a single, efficient network is a major advantage.
This new control strategy directly addresses one of the biggest hurdles to the widespread adoption of MVDC technology: the fear of instability. By providing a proven method to maintain voltage stability even during catastrophic load changes, the Tianjin team has removed a significant barrier. Utilities and grid operators can now be more confident in the safety and reliability of these systems.
Moreover, the economic benefits make the technology more attractive from a business perspective. An 11% reduction in operating costs is a powerful incentive for investment. This cost savings comes not just from better energy arbitrage with storage but also from reduced power losses in the grid and a lower need for expensive backup generation or emergency response measures.
The work also highlights the importance of a holistic approach to grid management. It’s not enough to have smart hardware or a smart software scheduler in isolation. True optimization comes from the tight integration of both. The long-term scheduler makes the big, strategic decisions for economic gain, while the short-term, hardware-based tan-droop control provides the rapid, automatic response needed for physical stability. This dual-layer approach creates a system that is both intelligent and robust.
Looking to the future, this research opens several new avenues. The team has already hinted at further work on optimizing the parameters of the tan-droop control itself. The “a” coefficient in their equation, which controls the steepness of the response, could be made adaptive, learning from system conditions to provide the perfect balance between a fast response and minimal overshoot. Furthermore, the same principles could be applied to larger, more complex grid architectures, including multi-voltage-level DC systems and even continental-scale supergrids.
In conclusion, the work by Xiao Qian, Lu Wenbiao, Jia Hongjie, Mu Yunfei, and Yu Xiaodan from Tianjin University represents a significant leap forward in the field of power systems. Their dual-time-scale hierarchical regulation and control method, built on an innovative tan-function-based droop control, offers a comprehensive solution to the twin challenges of economic efficiency and operational stability in modern MVDC grids. By proving that a system can be both highly economical and exceptionally resilient, they have provided a crucial blueprint for the next generation of power infrastructure. As the world moves toward a more electrified and renewable-powered future, technologies like this will be essential for building a grid that is not only green but also strong, smart, and dependable.
Xiao Qian, Lu Wenbiao, Jia Hongjie, Mu Yunfei, Yu Xiaodan, Key Laboratory of Smart Grid of the Ministry of Education (Tianjin University), Proceedings of the CSEE, DOI: 10.13334/j.0258-8013.pcsee.222632