Product1 publisher3 min readPublished
Sungrow triggered a wideband oscillation on purpose at a live 500 MW solar plant
Sungrow says it reproduced and suppressed wideband oscillations at a working 500 MW plant in Qinghai in September 2026, a test it describes as the first outside simulation, with grid-strength detection reported at 40 milliseconds.
The Product Desk · Product desk
What happened
- Sungrow says it reproduced and then suppressed wideband oscillations at a 500 MW solar-plus-storage plant, in what is described as the world's first field test of the technology.
- The test ran at a plant in Qinghai province, China, in September 2026, where engineers set off the fault on purpose.
- The company also says its grid-forming control strategy holds stable operation across a short-circuit ratio range of 1 to 40 and helps mitigate transient overvoltage.
Compiled by The Product DeskSomething wrong?How this is made
Why it matters
- capability A fault that had been too unpredictable to stage at an operating plant can now be summoned on demand. A buyer can ask to witness oscillation tolerance.
- constraint Only a detection time has been published, so an operator comparing inverter suppliers has no common figure for how fast the disturbance actually settles.
- exposure The inverter control parameters sit on both sides of this test, among the knobs that provoked the oscillation and inside the equipment that caught it, so the buyer takes the fix and the trigger from the same vendor.
- precedent Once one supplier has staged the fault at a 500 MW site, grid operators have a reason to ask the next one to do the same before signing a connection agreement.
Getting the fault to happen is the hard part. Sungrow's engineers established weak-grid conditions at the plant, which runs Sungrow PV inverters, and then worked three factors: the plant's power output, the system short-circuit ratio, and the inverter control parameters [8][9]. With those adjusted, a localized wideband oscillation appeared in a controlled environment, and the team measured how the inverters responded under different grid conditions [10]. Until now the unpredictable nature of these events had made them difficult to reproduce at operating plants [5].
The 40 milliseconds belongs to detection. Sungrow says its patented grid-strength adaptation technology identified the local grid strength within 40 ms, after which the equipment adjusted its control strategy to stabilize voltage and frequency [11]. Sungrow did not say how long the waveform took to settle after that change [20]. For scale on the detection window, the oscillations at Germany's BorWin1 HVDC project in 2014 ran between 250 and 350 hertz [14], which puts 40 ms at roughly 10 to 14 cycles of that kind of event [19]. BorWin1 damaged filter capacitors and contributed to a shutdown lasting around six months [14].
Pan Nian'an, chief engineer and chief expert at Sungrow, said: "For renewable energy to become a truly stable and reliable source of power, it is essential to develop autonomous fault self-healing and grid-support capabilities" [7].
What a plant does when a wideband oscillation arrives is come off the grid: these fluctuations can force generating equipment to disconnect or shut down entirely [4]. At Hornsea in 2019, subsynchronous oscillations were part of an event that caused widespread disconnections, with about 3.2 percent of UK system load lost and roughly a million users affected [15]. A buyer is trying to price that loss, and a detection time on its own is not yet a comparable number.
Separate the test from the spec sheet. The field test is a thing that happened at a named plant in a named month [1][3]. The claim that the grid-forming control strategy holds stable operation across a short-circuit ratio range of 1 to 40, and helps mitigate transient overvoltage, is the company's [13]. Sungrow also owns both halves of the demonstration: the inverter control parameters were one of the three knobs used to provoke the oscillation, and the adaptation logic that caught it sits in the same equipment [9][11].
Does this travel to your site? Ask whether the supplier can reproduce the oscillation using your output level and your short-circuit ratio at your point of connection, with a grid operator or lab watching, or only in a study. Then ask whether the number they quote is the time to identify the condition or the time to a settled waveform. Sungrow's published figure is the first one [11].
What to watch
- Whether a grid operator or independent lab publishes measurements from the Qinghai test.
- Whether any inverter supplier quotes a settling time in a connection test.
- Whether connection codes begin asking developers to reproduce an oscillation at the point of connection with a witness present.