Modular Pricing Strategy Boosts EV Charging Efficiency
In the fast-evolving landscape of electric mobility, a groundbreaking study has introduced a novel modular pricing framework designed to optimize charging behavior among electric vehicle (EV) users while balancing the interests of power suppliers, retailers, and consumers. The research, led by Xiao Bai from the Key Laboratory of Modern Power System Simulation and Control & Renewable Energy Technology at Northeast Electric Power University, presents a data-driven, customer-centric approach to electricity tariff design that could redefine how energy is priced and consumed in smart grid environments.
As global electrification accelerates and national carbon neutrality goals take center stage, the integration of EVs into power systems has become both an opportunity and a challenge. On one hand, EVs offer a pathway to decarbonize transportation; on the other, their charging patterns can strain grid infrastructure, particularly during peak hours. Traditional flat-rate or time-of-use pricing models often fail to account for the diverse behavioral patterns of EV owners, leading to suboptimal load distribution and missed opportunities for demand-side management.
To address these limitations, Xiao Bai and his team have developed a modular electricity price package system tailored specifically for EV users. Published in Power System Technology, the study proposes a two-stage clustering methodology combined with mass customization theory to create personalized, flexible, and responsive pricing structures. This innovation not only enhances user satisfaction but also strengthens grid resilience and improves economic outcomes across the energy value chain.
The foundation of this new pricing model lies in the accurate identification of user behavior patterns. Rather than treating all EV owners as a homogeneous group, the researchers employed a two-phase clustering technique to segment users based on their actual charging habits. The first phase utilized the Agglomerative Hierarchical Clustering (AGNES) algorithm, which excels at identifying natural groupings within datasets without requiring predefined cluster numbers. This was followed by K-means clustering, a widely used method known for its efficiency in refining cluster boundaries. By combining these algorithms, the team achieved a more robust and accurate classification of charging behaviors than either method could deliver alone.
Data from a business district in a northern Chinese city formed the basis of the analysis, with 96 data points collected daily at 15-minute intervals. From this, three distinct user profiles emerged: first-type users, primarily fleet operators who charge late at night; second-type users, private vehicle owners with home charging access who tend to charge in the early evening; and third-type users, urban dwellers reliant on public charging stations, often charging during morning hours. These behavioral archetypes provided the foundation for designing targeted pricing modules.
What sets this research apart is its application of mass customization principles—originally developed in manufacturing—to the service domain of energy retailing. Mass customization allows businesses to deliver individualized products or services at near-mass-production costs. In this context, the electricity price package is treated as a configurable product composed of interchangeable functional modules. Each module corresponds to a specific service feature, such as time-of-use flexibility, power level selection, vehicle-to-grid (V2G) participation, or reliability guarantees.
The modular structure includes six core components: charging power control, time-slot management, reliability options, V2G integration, ad-hoc charging reservation, and post-contract adjustment rights. These are not static offerings but dynamic building blocks that can be assembled based on user preferences and grid conditions. For instance, a user who prioritizes low cost and flexibility might receive a package emphasizing off-peak discounts and dynamic scheduling, while another focused on speed and convenience might be offered higher-power charging with guaranteed availability.
A critical innovation in the model is the correlation vector system used to match user needs with appropriate modules. Instead of relying on broad demographic assumptions, the framework calculates a comprehensive relevance score between individual charging demands and each available module. This is achieved through a weighted analysis of user priorities—such as cost sensitivity, charging speed, time flexibility, and willingness to participate in demand response programs—and how well each module satisfies those needs.
Through expert evaluation and field surveys, the team assigned weights to different service attributes for each user type. First-type users, mostly commercial operators, placed high value on cost efficiency and charging speed. Second-type users, typically homeowners, emphasized affordability and time flexibility. Third-type users, dependent on public infrastructure, showed greater responsiveness to incentives for shifting their charging times, making them ideal candidates for demand response initiatives.
Based on these insights, the system automatically selects a combination of mandatory and optional modules for each user segment. Mandatory modules—such as basic time-slot management and power control—are included in all packages to ensure minimum service quality. Optional modules, like V2G participation or premium reliability, are added based on individual relevance scores. This hybrid approach ensures both standardization and personalization, streamlining the configuration process while preserving customization depth.
Once the module set is determined, the next step involves attribute configuration—assigning specific values to each module’s parameters. For example, the time-slot management module can be set to “peak-only,” “off-peak only,” or “dynamic scheduling” depending on the user’s profile. Similarly, the V2G module can specify participation frequency (e.g., weekly, monthly, or random) and compensation rates. This level of granularity enables highly tailored packages that align closely with real-world usage patterns.
The final component of the framework is the pricing model, which integrates demand response dynamics and bilateral contracts between retailers and utilities. Unlike traditional pricing that focuses solely on cost recovery, this model considers the broader economic impact of user behavior changes. When users shift their charging to off-peak hours in response to price signals, they reduce grid congestion and lower generation costs for utilities. In return, retailers can negotiate lower wholesale prices through long-term contracts, passing some savings back to consumers while maintaining or even increasing their own margins.
The financial mechanics are carefully balanced. Retailers face reduced revenue from lower per-kWh sales prices but gain savings from reduced procurement costs and improved load factors. Consumers benefit from lower overall electricity bills despite potential price increases during peak periods, as their average cost per kilowatt-hour decreases. Utilities enjoy smoother load curves, reduced need for peaking plants, and enhanced system stability. The result is a true triple-win scenario that aligns economic incentives with operational efficiency and sustainability goals.
To validate the model, the researchers conducted a simulation using real-world load data. Three customized packages—EP1, EP2, and EP3—were designed for the three identified user types. The pricing structures reflected strategic trade-offs: higher peak rates to discourage overuse, competitive off-peak discounts to incentivize load shifting, and moderate mid-peak pricing to maintain flexibility. After implementation, the results were striking. First-type users increased off-peak charging by 11.93%, second-type users reduced peak demand by 12.96%, and third-type users shifted 10.91% of their peak load to lower-demand periods.
Aggregate load analysis revealed a significant flattening of the daily demand curve. Total energy consumption remained nearly unchanged, but the distribution improved dramatically: peak-period usage dropped from 11.33 MWh to 9.94 MWh, mid-peak usage rose slightly from 20.82 MWh to 21.27 MWh, and off-peak consumption increased from 22.62 MWh to 23.09 MWh. This shift not only reduced stress on the grid but also lowered the need for expensive peak-generation resources, translating into tangible cost savings.
From a financial perspective, the benefits were equally compelling. Users saw their total charging expenses decrease across all categories, despite the more complex pricing structure. First-type users saved over 400 yuan per day, second-type users saved nearly 656 yuan, and third-type users reduced their bills by more than 534 yuan. Meanwhile, the power retailer’s daily profit increased from 29,032.85 yuan under a flat-rate model to 30,287.70 yuan with the modular system, even after accounting for an additional 440.53 yuan in operational costs related to package management and marketing.
These outcomes underscore a fundamental shift in how energy services can be delivered. Rather than viewing pricing as a static regulatory tool, the study positions it as a dynamic, adaptive mechanism capable of shaping behavior, enhancing efficiency, and creating shared value. It also highlights the importance of data analytics and behavioral modeling in modern energy markets, where one-size-fits-all solutions are increasingly obsolete.
The implications extend beyond China’s evolving electricity market. As countries worldwide grapple with grid modernization, renewable integration, and transportation electrification, this modular approach offers a scalable blueprint for smarter energy pricing. Utilities in Europe, North America, and Asia facing similar challenges with EV adoption and peak load management could adapt this framework to local conditions, using regional load data and consumer surveys to calibrate the module weights and pricing parameters.
Moreover, the model’s flexibility makes it compatible with emerging technologies such as smart charging, bidirectional energy flows, and AI-driven energy management systems. As more EVs gain V2G capabilities, the inclusion of energy export modules could evolve into active grid support services, turning vehicles into mobile energy assets. In such a future, the modular pricing system could dynamically adjust compensation rates based on real-time grid needs, further enhancing its responsiveness and economic efficiency.
Another advantage of this approach is its scalability. Traditional tariff design often becomes unwieldy as the number of user segments grows. In contrast, the modular system maintains simplicity by reusing core components across different configurations. New user types can be accommodated by adjusting weights and adding specialized modules, without redesigning the entire pricing architecture. This reduces development time, lowers administrative overhead, and allows for rapid iteration in response to market changes.
Customer experience is also enhanced. Instead of being presented with a confusing array of fixed plans, users engage in a guided selection process that matches their lifestyle and priorities. The use of intuitive attribute settings—such as “charge overnight,” “fast charge when needed,” or “earn credits by delaying charging”—makes the system accessible even to non-technical consumers. This transparency builds trust and encourages active participation in demand response, a key factor in achieving long-term grid sustainability.
Regulators may also find value in this model. By promoting voluntary load shifting through market-based incentives rather than mandates, the system supports a more liberalized and consumer-oriented energy market. It aligns with global trends toward decentralized, participatory energy systems where users are not just passive consumers but active contributors to grid stability.
However, successful implementation requires more than just technical design. Stakeholder engagement, clear communication, and digital infrastructure are essential. Retailers must invest in user-friendly platforms that explain the benefits of each module and allow easy customization. Data privacy and security must be prioritized, especially when collecting detailed usage patterns. And policymakers should create enabling frameworks that support dynamic pricing and reward innovation in demand-side management.
Looking ahead, the research opens several avenues for further exploration. One direction is the integration of real-time pricing signals, allowing packages to adapt to day-ahead market conditions. Another is the inclusion of non-EV loads, such as home heating or battery storage, to create holistic household energy plans. Machine learning could also be used to continuously refine user segmentation and module relevance scores, making the system increasingly intelligent over time.
In conclusion, the modular electricity price package system developed by Xiao Bai and his team represents a significant advancement in energy economics and smart grid technology. By combining advanced clustering techniques, mass customization theory, and demand response modeling, the framework delivers a practical, scalable, and mutually beneficial solution to one of the most pressing challenges in the energy transition. It demonstrates that with the right design, pricing can be more than a financial tool—it can be a powerful lever for behavioral change, system optimization, and sustainable growth.
As the world moves toward a decarbonized future, innovations like this will play a crucial role in ensuring that the grid remains reliable, affordable, and resilient. The study not only contributes to academic knowledge but also provides actionable insights for industry practitioners and policymakers. Its success in a real-world setting underscores the potential of data-driven, user-centric approaches to transform the way we think about energy consumption.
Xiao Bai, Liu Jiatao, Yang Shiwei, Jiao Mingxi, Wang Daliang, Jiang Zhuo. Modular Design of Electricity Price Package for Electric Vehicle Users. Power System Technology, 2024, 48(11): 4544-4552. DOI: 10.13335/j.1000-3673.pst.2024.0527