The lithium-ion battery system, as a state-of-the-art electrical energy storage system, is the key component for hybridization. However, the performance of lithium-ion batteries degrades over time due to a number of internal secondary reactions and the fatigue of the electrode materials, as well as other mechanisms. By the end of its lifetime, the battery system can no longer satisfy the vehicle’s requirements due to its inadequate energy or power capability. Longevity is extremely important for large-scale battery systems—for example, those potentially integrated in plug-in hybrid (PHEV) commercial vehicles, where the cost of replacement is high and consistent driving performance throughout its lifetime needs to be guaranteed. Against this backdrop, FEV and the Institute for Combustion Engines of RWTH Aachen University (VKA) have developed a model-based method to optimize the operation strategy for plug-in hybrid commercial vehicles considering battery degradation during both driving and parking. This strategy includes techniques to design operating strategies for hybrid electric commercial vehicles that manage the trade-off between battery health and energy consumption costs, including those for both fuel and electrical energy.
Battery degradation has been widely investigated. Although some aging mechanisms are still not fully understood, it is found that the rate of aging is influenced by how the battery is used and stored. Several factors, such as high charge and discharge rates, deep depth-of-discharge, extreme state of charge (SoC), extreme temperatures, etc., are generally acknowledged to accelerate battery degradation. In vehicles, these factors are influenced by the on-board operation strategy, in which the usable SoC range and the engine turn-on thresholds are usually defined. However, imposing restrictions on the operation strategy to prolong battery life will most probably have a negative impact on the vehicle’s fuel economy.
Since energy cost and battery health depend on how the given PHEV is operated in everyday life, the engineers focused on the modeling of PHEV on-road power management as an important prerequisite for further optimization. In a second step, this vehicle model was coupled with a battery aging model that accounts for calendric and cyclic degradation and shows fuel and electrical energy consumption, as well as the battery’s state of health. For this, the research group from FEV and VKA adopted a so-called holistic aging model for Li(NiMnCo)O2-based lithium-ion batteries. This approach has been empirically identified as an accurate representation of health for an 18650 cell battery from Sanyo under both storage and cycle conditions. This model was developed based on results from accelerated aging tests examining different operational conditions.
Verified Data for Driving Pattern
The log representing the driving of a typical distribution truck from a German logistics company was used and analyzed to achieve a statistical view of commercial vehicles. The most frequent driving range could be defined from the histogram. This range is between 200 and 240 km, where the most frequent duration is between 9 and 9.5 h. Along with this information, the speed profile to solve the operation strategy problem was designed by combining two standard cycles for commercial vehicles (i.e. JE-05 Cycle and ‘World Harmonized Vehicle Cycle’ (WHVC)). The resulting model covers a distance of 237 km per day and a driving pattern consisting of 6 working days and 1 non-working day.
>> THE GOAL OF THE OPTIMIZATION WAS TO FIND AN OPERATION STRATEGY IN WHICH ENERGY CONSUMPTION COSTS ARE MINIMIZED WHILE THE BATTERY LIFETIME IS MAXIMIZED
In Search for the Optimal Operation Strategy
The goal of the optimization was to find an operation strategy in which energy consumption costs (i.e., fuel and grid electricity) are minimized while the battery lifetime is maximized. This optimization problem, however, has conflicting objectives. Since reducing the total energy costs requires higher battery usage, the battery generally tends to degrade faster. Therefore, a single optimal point does not exist; instead, a family of optimal solutions in the form of a Pareto front must be used.
Finding the Right Balance
To calculate the savings potential of an optimized operation strategy, a partially optimized strategy – without considering aging – was first generated as reference and a band of solutions was formed by NSGA-II. For electrical energy consumption greater than zero, a clear, linear trade-off between fuel consumption and electrical energy consumption was observed. This is the so-called Pareto frontier, which means that the truck cannot be made more fuel economic without consuming more electricity by changing the operation strategy. It indicates that reducing the fuel consumption by 1 l/100km will cause the consumption of an additional 4.8 kWh/100km extra electricity. For the points where no charging is needed, fuel consumption extends from 12.0 l/100km to 13.8 l/100km, which are results from the cases in which discharged energy is already compensated for by regenerative braking.
Plotting consumption in l/100km versus kWh/100km is a method of presenting the PHEV’s electrical and gasoline ‘fuel economy.’ Battery lifetime can be considered the third important variable and can be plotted on the 3rd axis of a 3D figure. After 35 generations of a population of 70 members, a Pareto optimum surface of optimal solutions is formed. This plot indicates that there exists a fundamental trade-off between battery lifetime and energy consumption costs. Namely, the lifetime can be prolonged from 5 to 15 years, but at the sacrifice of a 16% decrease in fuel economy and an 85% decrease in electricity consumption.
Calculating the Cost
To determine the most economic strategy, the prices of fuel, electricity, and the battery system must be taken into account. Since battery and energy prices are constantly evolving and difficult to forecast, average prices for diesel (1.365 €/l), battery (400 €/kWh) and electricity rates (0.275 €/kWh) are used. These costs represent the average prices in Germany from 2011 to 2015.
The optimal point was identified in the 2D trade-off results at a cost of 46.02 €/d including energy costs and battery wear, while that identified in the 3D trade-off results in 40.62 €/d. Savings up to 5.4 €/d (ca. 12%) could be achieved with an operation strategy that also takes battery wear into consideration. The battery life could be prolonged from 4.1 to 15.7 years. Comparing the parameters of both strategies, the maximum SoC, the available SoC range, and the engine turn-on speed for the 3D trade-off were significantly reduced, which extended the battery life.
The method developed for designing an optimized operation strategy for commercial vehicles will comprehensively reduce operation costs.
Comparisons of different solutions using the Pareto front confirm that imposing restrictions on the heuristic operating strategy – such as the usable SoC range and engine turn-on threshold in order to extend battery lifetime – yields higher total energy consumption. Given the cost profile of energy and batteries, the optimized operation strategy not only imposes a small available SoC window, but also preserves the battery at low SoC, which has proven to minimize calendric aging.
This approach can also be extended to other vehicle types, other battery degradation models, and other battery sizes. Emissions and drivability are also important factors in the operation strategy, and could be considered in future work. Furthermore, future work could include the optimal dimensioning of other powertrain components based on battery usage.