3 Charging Infrastructure Gaps China Is Closing—Fast
As China accelerates its transition to electric mobility, a critical bottleneck has emerged—not in battery technology or vehicle design, but in the strategic placement of charging infrastructure. While the nation leads the world in electric vehicle (EV) adoption, with over 20 million units on the road as of 2025, the mismatch between charger distribution and actual user demand continues to undermine grid stability, driver satisfaction, and renewable energy integration. A new methodology developed by researchers at Guangzhou Power Supply Bureau and South China University of Technology promises to reshape how cities plan for EV charging—not by adding more stations, but by placing them smarter.
At the heart of this breakthrough is a technique called the “plane demand method,” a departure from traditional point-based or traffic-intercept models that treat charging demand as isolated spikes. Instead, the plane demand approach views entire urban zones as continuous fields of energy need, dynamically shaped by vehicle density, traffic flow, and—critically—the location of distributed renewable sources like rooftop solar and small-scale wind. By integrating these layers into a unified spatial optimization framework, the team demonstrates a 99.7% reduction in charging overload risk, a 63.2% improvement in power stability, and a measurable 3.2% boost in user satisfaction—all with just one additional charger in a test network of 33.
For global investors, urban planners, and automotive executives watching China’s EV ecosystem, this isn’t just an academic exercise. It’s a blueprint for scaling sustainable mobility without overbuilding or destabilizing the grid. And as Western cities grapple with their own charging deserts and grid congestion, the implications are increasingly transnational.
The problem with most EV charging networks today is their reactive design. Planners often install chargers where real estate is cheap or where early adopters clustered—suburban shopping malls, downtown parking garages, highway rest stops. But as EV ownership spreads beyond tech-savvy elites into taxi fleets, logistics vans, and mass-market commuters, those assumptions collapse. A delivery van in Guangzhou’s industrial Baiyun District doesn’t need the same charging profile as a private sedan in Tianhe’s high-rises. Yet legacy planning treats them identically.
This mismatch creates three cascading failures. First, localized overloads: when dozens of EVs converge on a single station during peak hours, transformers trip, queues form, and drivers abandon trips. Second, underutilized assets: remote or poorly timed chargers sit idle, dragging down return on investment. Third—and most insidious—is the missed opportunity to align EV charging with renewable generation. Solar peaks at noon; if chargers are clustered in residential zones used mostly at night, that clean energy goes to waste or is curtailed.
The plane demand method tackles all three by redefining the planning unit. Instead of asking “Where should we put the next charger?” it asks, “How should we partition the city so that each zone balances load, access, and green power?”
The process begins with granular data: 15-minute intervals of charging load across existing stations, GPS-derived traffic patterns, and real-time output from distributed energy resources (DERs). Using this, the team constructs a multi-dimensional score for every potential zone, evaluating three key metrics: charging busyness, peak-valley aggregation, and DER alignment.
Charging busyness measures how close a zone’s current load is to the ideal 80% utilization threshold—high enough for efficiency, low enough to avoid stress. Peak-valley aggregation assesses whether charging demand within a zone naturally smooths out over time (e.g., some users charge midday, others overnight), reducing the need for grid ramping. DER alignment quantifies how well local renewable generation can offset charging demand, minimizing curtailment and carbon intensity.
Once the city is partitioned into balanced zones—six in the Guangzhou test case—the algorithm uses a Voronoi diagram to determine the optimal location for a new charger within each zone. Unlike simple centroid placement, this geometric method ensures every point in the zone is closer to its designated charger than to any other, minimizing average travel distance. Crucially, the model caps new installations per zone (one in the simulation) to enforce cost discipline, proving that strategic placement beats brute-force expansion.
The results are striking. In the baseline scenario—no new chargers—the network averaged high load with sharp peaks, especially between 8 p.m. and midnight. Randomly adding a charger reduced average load by just 2.95%. But the plane demand approach cut it by 23.97%, flattening the curve and eliminating dangerous overloads. The “danger index,” a composite of time and magnitude of overload events, plummeted from 896.3 to just 2.7.
Even more telling is the impact on power stability. In Zone 2—one of the most volatile—the variance in charging load dropped from 31,361 kW² to 12,403 kW² after optimization, lifting its stability score from 0.15 to 0.42 on a normalized scale. Across all zones, stability improved by over 63%, meaning fewer voltage sags, less equipment wear, and lower operational risk for utilities.
For drivers, the payoff is shorter trips and fewer waits. The model calculates user satisfaction based on a weighted combination of charging cost and travel distance. After optimization, satisfaction rose uniformly across zones—most notably in Zone 2, where it climbed from 1.16 to 1.20. While a 3.2% gain may seem modest, in a city with millions of daily EV trips, it translates to tens of thousands of saved hours and reduced range anxiety.
What makes this work particularly relevant to international audiences is its embedded response to China’s “dual carbon” goals—peaking emissions before 2030 and achieving carbon neutrality by 2060. Unlike Western models that treat EVs as standalone transport assets, China’s grid-planning culture increasingly demands source-load coordination: the tight coupling of generation and consumption in space and time.
The plane demand method operationalizes this principle. By factoring DER output directly into the zoning algorithm, it ensures new chargers are sited where solar or wind can directly serve them—turning EVs into mobile batteries that absorb surplus renewable energy rather than strain fossil-fueled peaker plants. This isn’t just efficiency; it’s systemic decarbonization.
For automakers like Tesla, BYD, or Volkswagen, whose charging networks are expanding globally, the lesson is clear: future-proof infrastructure must be co-designed with the grid, not bolted onto it. For investors in charging startups—many of whom have burned cash on underused hardware—the model offers a capital-efficient path to scale. And for policymakers in California, Germany, or Australia wrestling with grid integration, it provides a replicable framework that prioritizes intelligence over inventory.
Critically, the method avoids two common pitfalls of AI-driven planning. First, it doesn’t require perfect data. While GPS traces and smart meter readings enhance accuracy, the core algorithm works with aggregated traffic counts and historical load curves—data most utilities already collect. Second, it’s interpretable. Planners can see why a zone was drawn a certain way and adjust weights (e.g., prioritize DER alignment over travel distance in solar-rich regions).
The research team acknowledges limitations. The current model fixes charger power at a single value (100–630 kW range tested), ignoring cost differences between slow and ultra-fast units. Future work will layer in economic optimization—essentially turning the problem into a two-stage decision: where to build, and what type to install. They also plan to incorporate dynamic pricing signals, allowing the system to shift demand in real time rather than just shape infrastructure statically.
But even in its current form, the plane demand method represents a paradigm shift. It moves EV infrastructure planning from a siloed transportation issue to a core component of urban energy architecture. In doing so, it aligns three often-competing priorities: user convenience, grid resilience, and climate action.
As China rolls out its next-generation “smart charging” standards in 2026, expect techniques like this to move from academic papers into municipal codes. Cities like Shenzhen and Hangzhou are already piloting similar zone-based approaches. The ripple effects will be felt far beyond China’s borders—not just in exported EVs, but in exported planning logic.
For the global auto industry, the message is unambiguous: the race is no longer just about who builds the best battery. It’s about who builds the smartest ecosystem around it.
GUAN Junle, GAO Yuan, CHEN Fanghua, HU Shengqian, DENG Ming, LIANG Weiqiang, ZHANG Yongjun. The Location Optimization of Charging Facilities Based on the Plane Demand Method. Guangdong Electric Power, 2024, 37(10): 38–45. doi:10.3969/j.issn.1007-290X.2024.10.004