Sunwoda's 15C Flash Charging: What It Means for XPeng

Sunwoda (欣旺达) used a battery conference in Zaozhuang on 17 September to launch a 15C flash-charging battery system and an open "PV-storage flash-charging" ecosystem platform — and, more significantly, to announce that it is moving from being a cell supplier into building and investing in charging infrastructure. 张抗抗KK summarised the five things that matter. The numbers are extraordinary; the interesting question is what they mean for XPeng, which has a stake in the cell maker and a megawatt charging plan of its own.

What Sunwoda announced

1. The cell. Peak 15C. At room temperature, 10% to 70% in five minutes, and 10% to 97% in nine. Cell energy density is 190 Wh/kg, with system volume utilisation of at least 80%, and Sunwoda says it supports vehicles with more than 700 km of range.

2. The charger. A charging cabinet rated to 2.4 MW, with a single gun delivering up to 1.5 MW.

3. The answer to weak grid connections. Rather than upgrading the local supply, Sunwoda keeps the existing site and transformer and adds large-scale storage — in three tiers: 176 kWh / 0.7 MW for retrofitting old stations, 352 kWh / 1.4 MW for city stations, and 528 kWh / 2.1 MW for highway stations with multiple guns.

4. Four ways to pay for it. Selling components, delivering a turnkey solution, co-building the network, or joint investment and operation. The platform is initiated by Sunwoda.

5. The target. 10,000 megawatt-class flash-charging stations in China by the end of 2027, split between self-built and partner-built. As the post notes, that is a little over a year away.

Sunwoda also attached commitments that matter more than the top-line numbers: the system uses smart cells, an energy-domain controller and cloud management to identify more than 20 fault types with 2–10 days of advance warning at a claimed 99.77% accuracy, and non-commercial vehicles get unlimited ultra-fast charging events with a 10-year/300,000 km warranty.

Why 15C is a different conversation from 5C

The current generation of Chinese fast-charging marketing is built around 5C — XPeng's G9L being the obvious example, with its 800V architecture adding 450 km of range in 9 minutes and a 619.6 kW peak at roughly 727 V and 844 A.

Sunwoda's claim is not a marginal improvement on that. A 10–97% charge in nine minutes is a different shape of session: the interesting part is not the peak but that it holds to near-full. If it arrives in production cars at anywhere near these figures, "5C" stops being the frontier and becomes table stakes.

The implications for XPeng

1. XPeng has a seat at the table — but not an exclusive one. He Xiaopeng confirmed after the G9L launch that XPeng invested in three cell companies, and public reporting identifies them as CALB, Sunwoda and SVOLT, with their cells progressively entering XPeng's supply system. So a 15C cell from Sunwoda is, in principle, available to XPeng. It is equally available to everyone else — and Li Auto's relationship with Sunwoda is deeper and quantified: a ¥2.65 billion stake in Sunwoda Power, with the new-generation Li L8 switching entirely to Sunwoda cells. XPeng's advantage here is a position in the queue, not a lock on the technology.

2. XPeng's pack self-lead is exactly the layer that decides whether 15C is usable. From 2026 XPeng leads pack design, BMS and thermal management in-house. That matters more, not less, as charge rates climb: a 15C cell is only a 15C car if the pack's cooling can pull the heat out and the BMS can manage the rate against real operating conditions. XPeng has spent the last month arguing precisely this — that the pack, not the cell, is where a carmaker earns its charging advantage.

3. The station side is now running ahead of the car side. This is the least-discussed part of the announcement. A 1.5 MW gun cannot be exploited by an 800V car: 1.5 MW at 800 V implies around 1,875 A, well beyond what conventional CCS hardware delivers, which is why megawatt-class standards such as MCS push voltage up towards the 1,250 V class instead. An 800V platform is realistically a 600–800 kW platform — which is where the G9L already sits. So Sunwoda's announcement is, in effect, a specification for XPeng's next electrical architecture. The cell may arrive first; the car has to follow with a higher-voltage platform to use it.

4. The station model is a validation of X-Energy — and a potential substitute for it. XPeng's own megawatt station, built with Halo in Hong Kong, runs 1 MW peak, 1,000 A, four bays and a 400 kWh on-site battery, with the grid connection held to at most 240 kW so that no distribution upgrade is needed — the same architectural idea as Sunwoda's storage tiers, at a smaller scale. Sunwoda is now proposing to do that at 10,000 stations. Two readings follow: XPeng's approach is technically vindicated, and XPeng may not have to fund a megawatt network alone — it can plug into someone else's open platform, or use Sunwoda's four-part business model as a template for its own 1,000+ station, 13-country European plan, which is capital-heavy and slow to build alone.

5. The warranty is the quiet provocation. Sunwoda offering unlimited ultra-fast charging with a 10-year/300,000 km warranty on non-commercial vehicles is a direct answer to the question every buyer asks about fast charging — the same question XPeng's own powertrain chief addressed with data showing roughly a one-percentage-point difference in annual degradation between ultra-fast and slow charging. When a cell maker is willing to underwrite unlimited fast charging, that becomes the new baseline for the category, and it puts pressure on both cell suppliers and carmakers to match it.

What to watch

• Whether the 10,000-station target survives contact with construction schedules. Fifteen months is a short time for a network an order of magnitude larger than XPeng's own European plan; the post's own advice is to watch the build rate rather than the announcement.
• Which XPeng model first names a 15C cell — and whether it arrives with a voltage step above 800 V, which is the real gating factor.
• Whether XPeng keeps building its own network or shifts towards partner platforms, given that a supplier-initiated, open, storage-backed megawatt network is now on the table.

Caveats

Peak C-rates and peak gun power are laboratory and demonstration figures; the 10–97% time is quoted at room temperature, and no production car today can accept 15C. Station counts, storage tiers and business models are announcements, not installed infrastructure. And the "implications for XPeng" above are analysis, not statements from either company.

Source: 张抗抗KK on Weibo, 19 September 2026 — translated and adapted for XPENGTech. Battery-system detail via News18a's report on the Sunwoda launch; XPeng's pack-level strategy via he Xiaopeng's post-launch comments and News18a's technical explainer.