In 2025, at the engineering site, day and night began to lose their meaning. Our team worked through an intense week of commissioning, and at the hardest point some of us went three consecutive days without sleep. Every problem had to be located, discussed and tested again within a narrow engineering window. When the converter finally operated as expected, the moment lasted only a short time. The road to it had taken ten years.
The paper presents a 120 kV/360 MW hybrid commutated converter, or HCC. The figures describe its devices, operating principles and test results. What they cannot show is how the work moved from an idea that had no ready-made device, through years of laboratory and factory testing, to a valve tower connected to a real power system. That is the story behind our paper.
Figure 1. The ten-year path from the first HCC idea in 2017 to engineering deployment in 2025.
2017: an idea before the necessary device existed
The idea emerged in 2017 from a familiar weakness of long-distance direct-current transmission. Conventional thyristor converters are efficient, economical and able to carry enormous power, but their switches cannot actively turn off current. When the alternating-current grid is disturbed, the transfer of current from one phase to the next can fail, abruptly interrupting power transmission.
We asked a deceptively simple question: could a converter keep the strengths of the thyristor system while gaining the ability to turn off current when the grid could no longer complete the transfer? At the time, this was not a product-development task. The required reverse-blocking device, hybrid circuit and system-level control did not exist as a package that we could simply assemble. The question had to be broken into smaller questions, and each answer had to make the next step possible.
2020–2021: turning the idea into a device and a prototype
From 2020 to 2021, the work became physical. The team developed the reverse-blocking power devices needed for the concept and built a laboratory prototype around them. For the first time, the idea could be connected to cables, triggered by a control signal and judged by a measured waveform rather than a diagram.
The laboratory also exposed details that the original concept had not anticipated. When the controllable device turned off, a small charging current could continue flowing through the thyristor connected in series with it. Compared with the main current, it looked insignificant; electrically, it was enough to keep the thyristor conducting and prevent the hybrid connection from blocking voltage. The solution was a nonlinear metal-oxide varistor that allowed the transient current to pass and then rapidly suppressed it.
That episode changed the team’s understanding of the project. A system-level idea could succeed or fail because of a current lasting only microseconds. Device physics, auxiliary circuits, control and measurement could no longer be treated as separate subjects. The prototype was not merely a smaller converter; it was where we began learning how to work as one system team.
Figure 2. From a reverse-blocking power device to the laboratory prototype: in 2020–2021, the idea first became hardware that could be wired, triggered and measured.
2022–2023: leaving the laboratory for the factory test hall
A laboratory prototype can be adjusted between experiments. Engineering equipment must pass defined tests repeatedly and safely. In 2022 and 2023, the centre of the project therefore moved to the experimental factory of a converter-valve manufacturer, where we carried out two years of engineering-oriented type tests.
The scale changed the rhythm of the research. A test involved high-power equipment, protection settings, cooling, measurement systems and coordination among many people. If a waveform was abnormal, the next step was not simply to run the experiment again. We first had to understand what had happened, decide whether the cause lay in a device, circuit, control setting or measurement channel, and only then prepare the next test.
The valve eventually carried thousands of amperes and completed repeated controlled turn-offs under demanding conditions. More importantly, the team learned to read the same evidence together. Researchers who had once focused on a single device were now discussing the behaviour of an entire valve; colleagues working on system control learned to trace a result back to events inside a semiconductor chip. The equipment was being scaled up, but so was our shared judgement.
Figure 3. The engineering prototype and test platform in the converter-valve manufacturer's factory hall. Two years of testing moved the work from individual devices to a complete valve.
2024: changing the question from ‘can it work?’ to ‘can it be trusted?’
By 2024, demonstrating one successful operation was no longer enough. The question had become whether the technology could be trusted as power-system equipment. Reliability had to be argued layer by layer—from the reverse-blocking device and its auxiliary circuit to the valve structure, cooling system, control and protection.
This stage was less visible than a dramatic high-current test, but it was essential. Possible failure modes had to be examined, evidence had to be made consistent across different tests, and doubts from experienced engineers had to be answered with data. The team learned that engineering confidence is not created by one impressive result. It is accumulated through many ordinary results that agree with one another.
2025: the final distance to the grid
In 2025, the HCC moved into engineering application. The 120 kV/360 MW system was integrated into the Lingbao high-voltage direct-current project in Henan, China, and the technology is also being adopted for the Mengxi ultra-high-voltage direct-current project in Inner Mongolia.
The final week at the site compressed years of work into a succession of immediate decisions. Commissioning problems did not arrive in the tidy order of a research plan. They appeared in the system as a whole, and the team had to bring together everything learned since 2017. During the hardest stretch, work continued for three days without sleep. We remember the exhaustion, but even more clearly the dependence on one another: nobody at the site held the complete answer alone.
When the system passed its engineering tests, it was tempting to see the valve tower as the result. In reality, the result was also the team that had grown around it. We began with an idea divided among specialties. We reached the grid as a group able to move between the chip, the circuit, the valve and the power system.
Figure 4. The HCC valve tower at the engineering site. Ten years of work ultimately converged in equipment ready to operate in a real power system.
What the paper leaves between the lines
A paper compresses ten years into figures, methods and conclusions. That compression is necessary, but it hides the repeated experiments, factory days, reliability discussions and nights at the engineering site. It also hides how often progress consisted not of a breakthrough, but of finding the next problem clearly enough to solve it.
The technology still has further to go, including grid-forming control and greater modularity. But the experience has already changed how we understand research. Moving from an idea to infrastructure is not a single leap. It is a chain of evidence—and a team that grows strong enough to carry that chain from one scale to the next.
Acknowledgements: This work was supported by the Integration Projects of the National Natural Science Foundation of China-State Grid Joint Fund for Smart Grid (U2166602) and the National Natural Science Foundation of China (52241701 and 52407211).