Impact of Interdisciplinary Research on Planning, Running, and Managing Electromobility as a Smart Grid Extension

Alfredo D'Elia, Fabio Viola, Federico Montori, Marco Di Felice, Luca Bedogni, Luciano Bononi, Alberto Borghetti, Paolo Azzoni, Paolo Bellavista, Daniele Tarchi, Randolf Mock, Tullio Salmon CinottiIEEE Access 2015 · OpenReview

This paper, published in IEEE Access, provides a comprehensive look at how we can integrate electric vehicles (EVs) into the urban fabric not just as cars, but as active extensions of the Smart Grid. By combining traffic engineering, power systems analysis, and IoT communication, the authors present a holistic framework for planning and managing the "electromobility" ecosystem.

TL;DR

The paper proposes an interdisciplinary framework that treats electromobility as a core component of the Smart Grid, using co-simulation and semantic middleware to bridge the gap between transportation and energy. It demonstrates that managing EVs effectively requires a "multi-Internet" approach—integrating the Internets of Energy, Things, and Information—to ensure grid stability and user satisfaction.

What Problem Does This Paper Address?

As cities transition to electric mobility, two massive infrastructure systems—the power grid and the transportation network—are becoming inextricably linked. Traditionally, these are managed in silos. However, a traffic jam can delay EV charging, and a surge in charging demand can destabilize a local power feeder.

The multi-stakeholder ecosystem of a modern municipal utility, integrating power, transport, and telecommunications.

The authors tackle the "anxiety" of the EV era: for the driver, it’s the fear of running out of juice (range anxiety); for the grid operator, it’s the fear of localized blackouts caused by uncoordinated charging. The challenge is to create a system where heterogeneous devices (cars, chargers, sensors) from different manufacturers can communicate and cooperate in real-time to balance these competing needs.

Methods

The authors employ a multi-layered methodology to solve these integration challenges:

  1. Co-Simulation Environment: They integrated three distinct simulators to model the feedback loop between different domains:
  2. SUMO: For microscopic road traffic modeling.
  3. EMTP-RV: For high-fidelity electromagnetic transient power grid simulation.
  4. Riverbed Modeler: To simulate the 3G/UMTS communication networks that link the two.
The co-simulation architecture linking traffic (SUMO), power networks (EMTP-RV), and communication layers.
  1. Semantic Interoperability (Smart-M3): To handle the "Tower of Babel" problem where different devices use different data formats, they used a semantic publish-subscribe middleware. This allows devices to share information via a common "ontology" (a shared vocabulary of concepts).
  2. SDN for IoT: They proposed a Software Defined Networking (SDN) controller called MINA. This optimizes how data flows through the network, ensuring that critical grid control messages aren't delayed by less important traffic like infotainment streaming.
  3. Service Platforms: They developed three user-facing services: a Reservation Service to balance charger load, a Route Planner with range prediction, and a "What-If" Service to help internal combustion engine (ICE) drivers visualize how an EV would fit their daily habits.

Datasets and Validation

The study validates its models using several realistic data sources:

  • Geographic Data: OpenStreetMap (OSM) was used to import realistic road networks for the downtown area of Bologna and the wider Emilia-Romagna region in Italy.
  • Power Profiles: Realistic city load profiles (sampled every 15 minutes) and EV battery discharge models based on speed and acceleration.
  • Communication Profiles: Mobile traffic profiles based on 3GPP technical reports (TR 36.822).
  • Case Study: The paper highlights the "Manheim-neu" project in Germany, a village being rebuilt from scratch as a smart grid laboratory.

Key Insights

  • The Communication-Grid Feedback Loop: One of the most striking insights is how communication delays (latency) directly impact grid "fairness." If a control signal is delayed or lost, two charging stations on the same power line might receive different power limits, leading to an unequal distribution of energy among users.
  • Reservation as a Grid Tool: A reservation system isn't just a convenience for the driver; it’s a predictive tool for the grid. By knowing when and where a car will plug in, the grid can pre-allocate resources, preventing the "seeking" behavior that leads to EVs stranded with empty batteries.
  • Semantic "Glue": The authors argue that the key to a Smart Grid isn't just more sensors, but semantic interoperability. By using RDF triples (subject-predicate-object), the system can integrate a Siemens charger, a Nissan car, and an Enel grid controller without custom-coded drivers for every combination.

Key Results

  • Congestion Mitigation: The distributed control algorithm successfully prevented overcurrent conditions in 15 kV feeders by dynamically adjusting EV charging rates in response to real-time grid stress.
  • SDN Efficiency: The proposed SDN scheduling algorithm improved file-sharing throughput by 8% and reduced video jitter by up to 67% compared to standard "bin packing" or "load balance" methods.
  • User Success: In large-scale simulations of the Emilia-Romagna region, the Route Planning service significantly increased the "Success Probability" (the chance of reaching a destination without running out of charge) compared to traditional shortest-path routing.

Limitations and Future Work

While the framework is robust, the authors acknowledge several areas for growth:

  • V2G Integration: The current study focuses on "G2V" (Grid-to-Vehicle). Future work will explore Vehicle-to-Grid (V2G), where EVs act as mobile batteries to provide regulation services back to the utility.
  • Economic Modeling: The paper suggests more research is needed into variable pricing for reservations based on real-time grid limitations.
  • Scalability: While the co-simulation works for city sectors, scaling these high-fidelity electromagnetic simulations to a national level remains a computational challenge.

Summary updated

Abstract

The smart grid is concerned with energy efficiency and with the environment, being a countermeasure against the territory devastations that may originate by the fossil fuel mining industry feeding the conventional power grids. This paper deals with the integration between the electromobility and the urban power distribution network in a smart grid framework, i.e., a multi-stakeholder and multi-Internet ecosystem (Internet of Information, Internet of Energy, and Internet of Things) with edge computing capabilities supported by cloud-level services and with clean mapping between the logical and physical entities involved and their stakeholders. In particular, this paper presents some of the results obtained by us in several European projects that refer to the development of a traffic and power network co-simulation tool for electro mobility planning, platforms for recharging services, and communication and service management architectures supporting interoperability and other qualities required for the implementation of the smart grid framework. For each contribution, this paper describes the inter-disciplinary characteristics of the proposed approaches.
BibTeX
@inproceedings{delia2015impact,
  author = {Alfredo D'Elia and Fabio Viola and Federico Montori and Marco Di Felice and Luca Bedogni and Luciano Bononi and Alberto Borghetti and Paolo Azzoni and Paolo Bellavista and Daniele Tarchi and Randolf Mock and Tullio Salmon Cinotti},
  title = {{Impact of Interdisciplinary Research on Planning, Running, and Managing Electromobility as a Smart Grid Extension}},
  booktitle = {IEEE Access},
  year = {2015},
  url = {https://openreview.net/forum?id=QKdhbqGgv3},
}
Figure