With the rapid growth in electric vehicle penetration, the large-scale integration of electric vehicles poses significant challenges to grid stability, particularly in regions like China where the adoption of China EV is accelerating. To address this, I have developed a renewable energy system that combines energy storage technology with wind power, photovoltaic (PV) generation, and grid connections. This system aims to absorb highly volatile renewable energy while meeting the uncertain charging demands of electric vehicles. In this study, I employ a Monte Carlo sampling method to characterize the disordered charging behavior of electric vehicles and couple MATLAB with TRNSYS to build a dynamic simulation model. I propose an evaluation framework to assess system performance across different time scales, technical configurations, and regulation strategies. The analysis focuses on three key aspects: energy matching, flexibility, and environmental benefits, with multiple scenarios explored to derive optimal strategies for integrating electric vehicles into renewable energy systems.
The core of this research lies in modeling the stochastic nature of electric vehicle charging, which is critical for accurately evaluating system performance. The charging power for electric vehicles is simplified to a constant rate of approximately 3.5 kW for Level 1 slow charging, ignoring transient effects at the start and end of charging cycles. The start time of charging is modeled using a normal distribution, where the probability density function is given by:
$$f(x) = \begin{cases}
\frac{1}{\sigma_s \sqrt{2\pi}} \exp\left[-\frac{(x – \mu_s)^2}{2\sigma_s^2}\right] & \text{for } \mu_s – 12 < x < 24 \\
\frac{1}{\sigma_s \sqrt{2\pi}} \exp\left[-\frac{(x + 24 – \mu_s)^2}{2\sigma_s^2}\right] & \text{for } 0 < x < \mu_s – 12
\end{cases}$$
Here, \(x\) represents the start time of charging, \(\mu_s = 17.6\) is the mean, and \(\sigma_s = 3.4\) is the standard deviation. The daily driving distance, which influences the charging duration, follows a log-normal distribution:
$$f(x) = \frac{1}{x \sigma_d \sqrt{2\pi}} \exp\left[-\frac{(\ln x – \mu_d)^2}{2\sigma_d^2}\right]$$
where \(x\) is the daily mileage, \(\mu_d = 3.2\), and \(\sigma_d = 0.88\). The battery capacity of electric vehicles ranges from 20 to 30 kWh, aligning with typical China EV specifications. For renewable energy generation, the PV array power is calculated as:
$$P_{PV} = n_{pv} S_{pv} GHI \eta_{pv} \eta_{INV} f_{pv}$$
where \(n_{pv}\) is the number of PV panels, \(S_{pv}\) is the area per panel, \(GHI\) is the horizontal irradiance, \(\eta_{pv}\) is the PV efficiency, \(\eta_{INV}\) is the inverter efficiency, and \(f_{pv}\) is the derating factor. Wind turbine power is modeled as:
$$P_{wt} = \begin{cases}
0 & v \leq v_{in} \text{ or } v \geq v_{out} \\
p_{wt,R} \frac{v – v_{in}}{v_{rate} – v_{in}} & v_{in} < v < v_{rate} \\
p_{wt,R} & v_{rate} \leq v < v_{out}
\end{cases}$$
with \(v_{in}\), \(v_{out}\), and \(v_{rate}\) representing cut-in, cut-out, and rated wind speeds, respectively, and \(p_{wt,R}\) as the rated power. The energy storage system (ESS), based on batteries, follows:
$$E_{ba}(t+1) = E_{ba}(t)(1 – \delta) + P_{cha} \eta_{cha} \Delta t – \frac{P_{dis} \Delta t}{\eta_{dis}}$$
where \(E_{ba}\) is the battery energy, \(\delta\) is the self-discharge rate, \(\eta_{cha}\) and \(\eta_{dis}\) are charging and discharging efficiencies, and \(P_{cha}\) and \(P_{dis}\) are the respective powers.
To evaluate system performance, I define several key metrics. For energy matching, the on-site energy fraction (OEF) and on-site energy matching (OEM) are used:
$$OEF = 1 – \frac{\int_{t_1}^{t_2} P_{grid-load}(t) dt}{\int_{t_1}^{t_2} P_{load}(t) dt}$$
$$OEM = 1 – \frac{\int_{t_1}^{t_2} P_{RE-grid}(t) dt}{\int_{t_1}^{t_2} P_{RE}(t) dt}$$
where \(P_{grid-load}\) is grid power supply, \(P_{load}\) is load power, and \(P_{RE-grid}\) is renewable energy exported to the grid. Flexibility is assessed via grid integration level (GIL) and net interaction level (NIL):
$$GIL = \frac{\int_{t_1}^{t_2} P_{grid-load}(t) dt}{\int_{t_1}^{t_2} P_{load}(t) dt}$$
$$NIL = \frac{\int_{t_1}^{t_2} P_{grid-load}(t) dt + \int_{t_1}^{t_2} P_{RE-grid}(t) dt}{\int_{t_1}^{t_2} P_{load}(t) dt}$$
Environmental benefits are measured by equivalent CO2 emissions (ECE):
$$ECE = CEF (E_{imp} – E_{exp})$$
with \(CEF\) as the carbon emission factor, \(E_{imp}\) as imported grid energy, and \(E_{exp}\) as exported energy.
The simulation is based on an office building in Shenyang, China, with a total area of 1025 m² and peak cooling and heating loads of 35.92 kW and 59.08 kW, respectively. The building integrates a heat pump system, and the renewable energy system includes PV arrays and wind turbines. The ESS consists of batteries with a capacity of 132.46 kWh. The electric vehicle load is modeled using Monte Carlo simulations in MATLAB, coupled with TRNSYS for dynamic system analysis. Key parameters for the building envelope and system components are summarized in the following tables.
| Component | Heat Transfer Coefficient [W/(m²K)] |
|---|---|
| Exterior Wall | 0.10 |
| Roof | 0.10 |
| Ground | 0.12 |
| Windows | 1.00 |
| Component | Specifications |
|---|---|
| PV Array | Peak voltage 40.9 V, peak current 13.45 A, 56 panels, peak power 550 W per panel |
| Wind Turbine | Rated power 20 kW |
| ESS | Battery capacity 132.46 kWh, storage volume 7 m³ |
| Heat Pump | Cooling capacity 19.76 kW, heating capacity 27.65 kW |
In the analysis of typical days, I examine energy flows under different seasons. For a transition season day, the system generates 263.88 kWh, meeting the load without grid support during most hours. The ESS maintains a state of charge (SOC) between 0.7 and 0.9, discharging in the evening to reduce grid dependence by 27.56 kWh. During a cooling season day, the load consumption is 223.59 kWh, while renewable generation is 160.34 kWh, leading to grid supplementation of 77.32 kWh. The ESS saves 48.60 kWh by shifting loads. In the heating season, similar patterns are observed, with ESS saving 35.48 kWh. These results highlight how the integration of ESS mitigates the impact of disordered electric vehicle charging, enhancing self-sufficiency.
For annual performance, I compare systems with and without ESS. With ESS, the average OEF increases by 48.20% to 62.20%, and OEM rises by 39.97% to 72.07%. Flexibility improves, with GIL decreasing by 34.87% to 37.79% and NIL by 37.77% to 65.84%. Environmental benefits are also evident, as ECE reductions of 6.59% are achieved. These metrics demonstrate that ESS significantly enhances the system’s ability to handle electric vehicle loads, particularly in seasons with high energy demands. The following table summarizes the annual comprehensive indicators.
| Indicator | With ESS | Without ESS | Improvement |
|---|---|---|---|
| OEF (%) | 62.20 | 41.99 | 48.20% |
| OEM (%) | 72.07 | 51.50 | 39.97% |
| GIL (%) | 37.79 | 58.06 | -34.87% |
| NIL (%) | 65.84 | 105.84 | -37.77% |
| ECE (kgCO2) | Reference | Higher | -6.59% |
I further analyze three regulation strategies: STR1 (priority to building and electric vehicle load), STR2 (priority to ESS charging), and STR3 (ordered charging with vehicle-to-grid capability). Under STR1, OEF averages 73%, while STR2 yields 63%. STR3 achieves the best balance, with OEF at 66.71%, OEM at 73.20%, GIL at 33.29%, and NIL at 52.63%. This strategy effectively utilizes electric vehicles as distributed storage, reducing grid interactions and enhancing renewable energy utilization. The ordered charging approach, in particular, aligns electric vehicle charging with renewable generation peaks, minimizing the disturbances caused by disordered charging behavior.
In conclusion, this study demonstrates that integrating energy storage and implementing ordered charging strategies can significantly improve the performance of renewable energy systems in the context of electric vehicle adoption. The use of Monte Carlo simulations for modeling electric vehicle behavior provides a realistic assessment of uncertainties. The findings are particularly relevant for China EV markets, where rapid growth necessitates robust grid integration solutions. Future work could explore real-time optimization algorithms and larger-scale deployments to further enhance system resilience and sustainability.