Real-time renewable energy incentive system for electric vehicles using prioritization and cryptocurrency
Introduction
The rising awareness of global warming and energy sustainability has led to a worldwide effort to introduce more Renewable Energy Resources (RES) into the current energy mix. Currently, the most promising and popular RES include Photovoltaics (PV), wind and geothermal [1], [2], [3]. However, due to their variability and intermittency issues, it’s not a trivial task to integrate these RES into the electric grid. For example, PV energy has created a significant increase in daytime generation and a valley of net demand in the afternoon, followed by another net demand peak at night, most famously described by California “duck curve” [4]. Wind energy suffers from problems such as generation curtailment due to transmission and demand issues all over the world [1]. Extensive work has been done to address the RES integration issues in microgrid [5], [6], [7], [8]. Dawound [5] reviewed common optimization techniques for microgrid with RES. Mehdizadeh [6], Javidsharif [7] and Lorenzo [8] propose scheduling of renewable-based microgrid using information gap decision theory, multi-objective optimization and model predictive control.
At the same time, the ownership of electric vehicles (EVs) has been rapidly growing all over the world. In United States, the projected sale of EVs is expected to reach 2.3 million, or 19% of the total sales by 2050 [9]. This creates an increase in electric energy demand and puts additional stress on the power grid. The problem will be further amplified with more powerful Direct Current Fast Chargers (DCFC) and vehicles with larger batteries. Uncontrolled EV charging creates large load variations in the local distribution grid, amplifying the peak further and reducing system reliability and power quality [10], [11], [12]. To understand how EV charging will impact the power distribution grid, Mozafar [13] studies the effects of EVs on power system demand, stability and reliability. Kheradmand [14] evaluates the distribution grid’s well-being and reliability in the presence of EVs. Amini [15] proposes a two-stage approach to allocate EV charging lots and RES in the distribution network. Mohammadi [16] presents a decentralized decision-making algorithm for collaborative optimal power flow in the transmission and distribution networks. Nienhueser [17] and Schuller [18] discuss the impact and future of RES integration with EVs considering EVs’ economic and environmental impacts and load flexibility.
With coordination and scheduling, however, EVs could be a valuable asset as a shiftable demand in the grid to alleviate over-generation and night peak problems [19]. Modern EVs have the ability to serve as a temporary controllable load and an aggregation of EVs can be a powerful tool to adjust the load variation on the electric grid without major impact on individual users. If the coordination of an EV aggregation is coupled with the dynamics of RES generation, the effect of EVs’ load on the external grid can be minimized while offsetting the negative impact of RES on the grid. There exists literature on how to guide electricity users’ behaviors using incentives, demand response [20], [21], [22], [23], [24], [25] and smart charging algorithms for aggregated EVs [26], [27], [28]. Valles [20], Eissa [22] and Yu [23] model incentive programs considering user responsiveness, multiple resources’ coordination and hierarchical electricity markets. Zhang [21], Haghi [24] and Zhao [25] provide market investigation, benefit analysis and execution experience of providing incentives. Rubino [26], Mortaz [27] and Lu [28] review and propose microgrid scheduling and dispatch algorithms with EV smart charging. Most of the current literature discusses traditional monetary incentive and few incentive programs on electric vehicles’ consumption behavior exist in current market.
Many works of literature have investigated approaches to increase the utilization of RES with smart EV charging. Mouli [29] designed a solar harvesting system used to charge EVs. Researchers [30], [31], [32], [33] consider EVs as a part of larger loads, such as households or buildings, supported by a wide variety of RES and use linear programming [30], convex programming [31], mixed-integer programming [32] and semi-Markov decision process (SMDP) [33] to find the optimal strategies to maximize RES utilization. Most of the technologies of RES maximization either involve Vehicle-to-Grid (V2G) capability or placement of energy storage. However, large energy storage is still expensive and the V2G technology is also far from being standardized and popularized. In fact, there are few EVs with V2G capability on the road today.
Smart charging algorithm designs with various optimization approaches are also widely proposed [34], [35], [36], [37]. Several real-time smart charging algorithms are also proposed based on use priorities [38], [39], [40]. Many minimize energy costs for drivers and aggregators assuming a dynamic or Time-of-Use (TOU) pricing scheme. While this is an effective approach, flat-rate pricing is still dominant in the real world and there are many obstacles for full implementation of dynamic pricing. If no dynamic pricing or monetary incentive is imposed, the performance improvement of cost reduction would be limited. Moreover, most of the studies assume complete information on user behavior and full control of each EV. This, however, is hardly the case in actual implementation. Experiments and implementation experiences of optimal energy management are reported [41], [42], [43], [44]. Peng [41] and Lopez [42] provide deployment experience of renewable microgrid in Singapore and Andean countries. Quashie [43] and Arcos [44] verify the effectiveness of their proposed control algorithms through experiments conducted in Canada and Spain. Zhang [45] reports an operating Energy Management System (EMS) and proposes a flat-rate real-time smart charging algorithm assigning priority based on users’ charging profile and demand. These reported systems are mostly multi-resource Demand Side Management systems without particular focus on EV and RES integration in microgrid.
The emerging technology, blockchain [46], shows an opportunity to provide incentives to users without going through the traditional pricing scheme posed by utility or aggregators. Blockchain provides decentralized and mutually trusted cryptocurrency transaction system for participants. Consensus algorithms, such as Proof of Work (PoW) and Proof of Stake (PoS), are used in blockchain to make sure data is verified and consistent. Different types of blockchains are designed in terms of who can participate in the consensus process. A summary of three types of blockchains is shown in Table 1 [47]. For public blockchain, anyone can participate in the consensus process to verify the written transactions and all data is open to public. However, due to the large pool of consensus nodes, the efficiency of a public blockchain can be low. For consortium and private blockchains, only a set of nodes are selected to participate in the consensus process. Therefore, the efficiency of the network can be higher. The transaction data can also be restricted to participating nodes, making it suitable for enterprises wishing to keep their data private. However, due to the small number of consensus participants, there is still chance for data on the blockchain to be tampered. So far, only a few efforts have been published on incorporating blockchain within the energy industry. Sikorski [48] presented a market design to implement a blockchain-based electricity market in the chemical industry. Mengelkamp [49] presented a case study of Brooklyn Microgrid implementing a local energy market with blockchain. While this blockchain market is active with trading activities, it is still “virtual” as it doesn’t change the actual power flow in the microgrid with the transaction. More practical use of blockchain with smart grid should be encouraged to reduce cost for users and aggregators and improve operation efficiency.
In this paper, we propose a smart EV charging system to motivate and incentivize users to use more RES collectively even on a flat-rate pricing system. The market design can provide monetary and non-monetary incentives while reducing operating and energy cost for aggregators. The system does not require energy storage and is suitable for large-scale adoption. The system is online and designed with consideration of practical implementation issues, setting itself apart from other advanced yet nearly impossible to implement algorithms. The algorithm is implemented on the campus of the University of California, Los Angeles (UCLA) to evaluate its effectiveness. The contributions and novelties of this work are as following:
- 1.
Overall, the paper proposes a novel system design that provides incentives to EV users to alleviate the problems of RES over-generation and mismatch between RES supply and demand. Real-time control algorithms for EV charging stations and optimal strategies for users to reach a Nash equilibrium local optima are also proposed. The algorithms incorporate considerations of real EV charging station implementation and are verified with a 15-month experiment.
- 2.
A scalable incentive design is proposed to guide user behaviors. The system implements the incentives without imposing Time-of-Use (TOU) pricing and using energy storage. The system incorporates non-monetary incentives (priority) and monetary incentives (blockchain-based cryptocurrency).
- 3.
A blockchain-based cryptocurrency trading framework is designed to provide monetary incentives to users to pass down the cost-savings of the aggregators. Unlike most cryptocurrencies that only work as an alternative currency, the proposed cryptocurrency has both the abilities to convert to fiat money and change the physical power flow, which is one of the first in the energy industry.
- 4.
Online priority ranking algorithms, such as Priority Sharing and Priority Round Robin, are proposed and optimized to facilitate practical implementations involving EVs and EV charging stations operating in SAE J1772 standard. User-level optimal strategies are also designed and proposed to help the system to achieve local optima, verified by numerical simulations.
- 5.
An experiment with workplace charging running Priority Round Robin for 15 months is implemented to evaluate the effectiveness of the algorithm on the UCLA campus. The test site has shown consistent local over-generation and demand-supply mismatch problems before the experiment. The experiment and data analysis using Welch’s t-test show that local solar consumption ratio has increased 37% with non-monetary incentives applied alone.
The rest of this paper is organized as follows.
Section 2 explains the system design and its market mechanism. Sections 3 Prioritization algorithm, 4 Optimal strategies show the fundamental prioritization algorithm and optimal strategies for incentivized drivers and aggregators. Section 5 shows the numerical simulation of the algorithm and Section 6 shows the experimental result of the algorithm, with discussion and conclusion in Sections 7 Discussion, 8 Conclusion.
Section snippets
Incentive system design
As commercial EV charging service providers do not have control over customer behaviors, such as when they arrive, leave or how much energy they request, it is essential to design an incentive system in which it is only reasonable for the user to follow the incentive guidance. Traditionally, this is achieved by posing different prices at different time. This section presents the system design to incentivize EV users to collectively adjust their charging time so that a better overall RES
Prioritization algorithm
The operations of EV charging stations implementing user prioritization is discussed in this section.
Optimal strategies
In this section, the optimal strategies at the aggregator and incentivized drivers’ levels are discussed for charging boxes running Priority Round Robin algorithm. The algorithms are online as they are executed with new status updates. The input of the algorithms is real-time information and each user’s own estimation of RES.
Numerical simulation
In this section, a numerical Monte Carlo simulation is implemented to show how users’ decision would change the system’s RES usage and whether the Priority Round Robin algorithm would effectively respond to incentivized users’ behavior change.
User interactions are simulated and repeated for 30 times to eliminate the random effects. There are a total of 4 users simulated as customers of the charging box and they arrive to charge every day with a charging box with four sub-stations. On each day,
Experimental results
If users are indeed incentivized by the priority system, they will optimize their action according to Algorithm 5, Algorithm 6 and system consumption will improve similar to Fig. 9, Fig. 10. If the incentive system fails, they will remain “uncontrolled”. In order to verify actual users’ response to such design, an experiment was implemented with one charging box and 15 EV long-term employee (joining at a different time) for 15 months in a workplace parking lot of University of California, Los
Discussion
Today, the trends for large battery EVs (including semi-trucks) and high power charging stations are clear. The electric grid will face larger challenges to balance and manage the load from EVs. The incentive mechanism presented in this paper can be adopted in such scenarios. However, in order to have a truly effective incentive and load management system, special attention of research and technology development should be paid in the following areas:
- 1.
Interoperability between devices: Currently
Conclusion
In this paper, we presented a scalable renewable energy maximization system, which incorporates electric vehicle charging dynamics, based on the concepts of user prioritization and blockchain cryptocurrency. The proposed system can not only reduce the transmission load of the distribution grids, which benefits the operators and customers, but also solve the over-generation problem of renewable energy, also known as the California duck curve. The proposed system and its novelties can be
Acknowledgement
This work has been sponsored in part by grants from the California Energy Commission (CEC) entitled “Demonstration of PEV Smart Charging and Storage Supporting Grid Operational Needs”. Sponsor Award number: EPC-14-056.
The authors would like to thank Xi Chen and Yishen Wang of GEIRI North America for their valuable advices and help.
References (58)
Wind power and externalities
Ecol Econ
(2017)- et al.
The prospects of the expanded diffusion of cogeneration to 2030–study on new value in cogeneration
Appl Therm Eng
(2017) - et al.
Risk-based energy management of renewable-based microgrid using information gap decision theory in the presence of peak load management
Appl Energy
(2018) - et al.
Multi-objective short-term scheduling of a renewable-based microgrid in the presence of tidal resources and storage devices
Appl Energy
(2018) - et al.
Power system impacts of electric vehicles in Germany: charging with coal or renewables?
Appl Energy
(2015) - et al.
The optimization of dc fast charging deployment in california
Appl Energy
(2015) - et al.
Optimal fast charging station placing and sizing
Appl Energy
(2014) - et al.
Innovative appraisement of smart grid operation considering large-scale integration of electric vehicles enabling v2g and g2v systems
Electr Power Syst Res
(2018) - et al.
Simultaneous allocation of electric vehicles parking lots and distributed renewable resources in smart power distribution networks
Sustain Cities Soc
(2017) - et al.
Economic and environmental impacts of providing renewable energy for electric vehicle charging – a choice experiment study
Appl Energy
(2016)