Eric Bibeau

PEV Charging Profile Prediction and Analysis Based on Vehicle Usage Data

Present-day urban vehicle usage data recorded on a per second basis over a one-year period using GPS devices installed in 76 representative vehicles in the city of Winnipeg, Canada, allow predicting the electric load profiles onto the grid as a function of time for future plug-in electric vehicles. For each parking occurrence, load profile predictions properly take into account important factors, including actual state-of-charge of the battery, parking duration, parking type, and vehicle powertrain. Thus, the deterministic simulations capture the time history of vehicle driving and parking patterns using an equivalent 10 000 urban driving and parking days for the city of Winnipeg. These deterministic results are then compared to stochastic methods that differ in their treatment of how they model vehicle driving and charging habits. The new stochastic method introduced in this study more accurately captures the relationship of vehicle departure, arrival, and travel time compared to two previously used stochastic methods. It outperforms previous stochastic methods, having the lowest error at 3.4% when compared to the deterministic method for an electric sedan with a 24-kWhr battery pack. For regions where vehicle usage data is not available to predict plug-in electric vehicle load, the proposed stochastic method is recommended. In addition, using a combination of home, work, and commercial changing locales, and Level 1 versus Level 2 charging rates, deterministic simulations for urban run-out-of-charge events vary by less than 4% for seven charging scenarios selected. Using the vehicle usage data, charging scenarios simulated have no significant effect on urban run-out-of-charge events when the battery size for the electric sedan is increased. These results contribute towards utilities achieve a more optimal cost balance between: 1) charging infrastructure; 2) power transmission upgrades; 3) vehicle battery size; and 4) the addition of new renewable generation to address new electric vehicle loads for addressing energy drivers.

Profile of Charging Load on the Grid Due to Plug-in Vehicles

The paper presents a study of the profile of the load imposed on a power system by grid-charging of the onboard battery pack of electric and plug-in hybrid vehicles. The study uses a large database of field-recorded driving cycles stamped with parking times and locations to predict realistic driving habits of drivers in an urban setting. A fuzzy-logic inference system is designed to emulate the decision-making process of a driver when deciding to charge the vehicles battery. The charging load is then estimated on an hourly basis for a number of electric and plug-in hybrid vehicles with different storage capacities. Level-1 and level-2 charging regimes as well as two scenarios for charging, namely charging at home and charging at home and work, are considered. The load profile is presented as an hourly probability of charging for each vehicle type. The results demonstrate how penetration of plug-in hybrid and electric vehicles affects the load on a utility network.

A study of bubble ebullition in forced-convective subcooled nucleate boiling at low pressure

Abstract An investigation of forced-convective subcooled nucleate boiling was carried out using high speed photography. Experiments were performed using a vertical circular annulus at atmospheric pressure, for mean flow velocities of 0.08–1.2 m s −1 and subcoolings of 10–60°C. The filmed conditions are defined relative to the onset of nucleate boiling and the onset of significant void. The following observations were made: (i) bubbles do not grow and collapse on the heated wall, but eject into the flow for subcoolings below 60°C; (ii) after the onset of nucleate boiling, bubbles slide away from the nucleation site and later eject into the flow; (iii) bubbles condense while sliding on the wall; and (iv) bubbles generated near the onset of nucleate boiling conditions slide for a distance of up to 50 mm, while for other conditions the total axial distance traversed by the bubbles is less than 2 mm on average. The maximum bubble diameter and condensation time are shown to be influenced by the location relative to the onset of significant void.

Wind Turbine Power Curve Modeling Using Advanced Parametric and Nonparametric Methods

Wind turbine power curve modeling is an important tool in turbine performance monitoring and power forecasting. There are several statistical techniques to fit the empirical power curve of a wind turbine, which can be classified into parametric and nonparametric methods. In this paper, we study four of these methods to estimate the wind turbine power curve. Polynomial regression is studied as the benchmark parametric model, and issues associated with this technique are discussed. We then introduce the locally weighted polynomial regression method, and show its advantages over the polynomial regression. Also, the spline regression method is examined to achieve more flexibility for fitting the power curve. Finally, we develop a penalized spline regression model to address the issues of choosing the number and location of knots in the spline regression. The performance of the presented methods is evaluated using two simulated data sets as well as an actual operational power data of a wind farm in North America.

Battery Storage Sizing in a Retrofitted Plug-in Hybrid Electric Vehicle

This paper develops a simulation-based framework for optimal sizing of the additional energy storage required to retrofit a hybrid electric vehicle (HEV) to a plug-in hybrid electric vehicle (PHEV). Simulations are conducted on a vehicular model developed for a midsize sedan (Toyota Prius) using a new weekly vehicle-usage profile constructed for average driving and most probable parking times based on the data collected in the city of Winnipeg (Canada). Three battery technologies that are commercially available for electric vehicle propulsion are used in the simulations to determine the optimal sizing of the battery storage, given the constraints on the volume of the battery pack for lowest cost. Overnight-charging and opportunity-charging scenarios are also implemented in the simulation, and their impact on the optimal sizing is discussed.

On Conversion of Hybrid Electric Vehicles to Plug-In

The retrofit conversion of currently available hybrid electric vehicles (HEVs) to plug-in HEVs (PHEVs) is studied in this paper through experiments and simulations using the powertrain system analysis toolkit (PSAT). First, a rule-based fuzzy controller of the battery energy-management unit is developed to simulate different energy-management policies. Second, by modifying the energy-control strategy, the model of the conversion PHEV (C-PHEV) is verified with experiments. Finally, the C-PHEV model is used to simulate different battery energy-management control strategies. The results show improvement in fuel economy, whereas the energy-management controller discharges the power through the plug-in battery pack only when the state of charge of the base vehicle battery is close to its minimum value. This method keeps the advantage of driving in electric mode using a combination of two batteries and optimizing the use of regenerative braking capabilities, which is the main advantage of HEVs. It is also found that increasing the power threshold of the internal combustion engine (ICE) improves the performance of C-PHEV. Increasing the ICE power threshold increases the engine efficiency by running the engine in its efficient points. It also drives the vehicle in electric mode in higher power demands.

DEVELOPMENT OF 3-D ANAEROBIC DIGESTER HEAT TRANSFER MODEL FOR COLD WEATHER APPLICATIONS

Anaerobic digester heat losses must be minimized to reduce heating requirements in cold weather applications. Digesters must be designed with proper insulation to control manure temperature through a variety of ambient weather conditions, but the additional insulation to mitigate cold weather operations must not impede digester economics. These design aspects are difficult to address for distributed power generation for on-farm anaerobic digestion, as these preclude the use of large-scale systems to reduce capital and operating costs. To investigate and address these issues, a 3-D mathematical model for simulating heat transfer for anaerobic digesters for cold weather conditions is developed and used to optimize the various geometrical parameters to achieve an adequate design that limits heat losses. An anaerobic digester heat transfer model based on computational fluid dynamics (Fluent 6.1) is used to calculate the heat transfer through the cover, floor, and walls of a below-ground lagoon-type digester. Simulated heat transfer results are compared to a 1-D numerical model and validated against experimental data using an operating plug-flow anaerobic digester. The 3-D predictions have the advantage of avoiding space-averaged boundary conditions and can account for conduction in all three directions in a digester. Simulated results agree reasonably well with the measurements and the one-dimensional model. Numerical simulations are performed for four digester configurations: (1) rectangular with arched top, (2) rectangular with flat top, (3) cylindrical with flat top, and (4) cylindrical with conical bottom. Sensitivity analysis has demonstrated the heat loss through the digester cover, floor, and walls for different geometrical dimensions. Comparisons of the total heat loss show that the cylindrical digester with a flat top offers the best geometry to minimize heat losses in cold weather applications, and that the heat-loss-to-biogas-heat ratio (HLB) is an important parameter for characterizing digester operations in cold climates.

Subcooled void growth mechanisms and prediction at low pressure and low velocity

Abstract Measurements were made of void growth and wall temperatures for two circular annular geometries and forced-convective subcooled nucleate boiling conditions, with flow rates between 0.02 and 0.20 kg/s, pressures of 1.05, 2 and 3 bar, and inlet temperatures between 30 and 90°C. Void growth results at low pressure show that the highly subcooled void region is important, and void fraction at the onset of significant void may be as high as 10%. Observations of bubble ebullition and detachment mechanisms obtained from a high speed photographic study were compared to assumptions generally used in void growth modeling. The photographic results show that bubbles do not travel far downstream after nucleating, that there is no region of attached void and that bubbles slide along the wall before being ejected into the flow after the onset of nucleate boiling. Unlike findings for high pressure systems reported in the literature, the onset of significant void was found to be independent of (i) bubble detachment, (ii) the location where the bubble is first ejected from a bubble layer and (iii) the transition from partial nucleate boiling to fully developed boiling. A phenomenological void growth model is presented which accounts for the vapor volume inside a heated channel at atmospheric pressure, and includes the bubble ebullition cycle, formulated on the basis of information obtained from the high speed photographic study.

Using Large Driving Record Samples and a Stochastic Approach for Real-World Driving Cycle Construction: Winnipeg Driving Cycle

The challenges in the development of plug-in electric vehicle PEV powertrains are efficient energy management and optimum energy storage, for which the role of driving cycles that represent driver behaviour is instrumental. Discrepancies between standard driving cycles and real driving behaviour stem from insufficient data collection, inaccurate cycle construction methodology, and variations because of geography. In this study, we tackle the first issue by using the collected data from real-world driving of a fleet of 76 cars for more than one year in the city of Winnipeg Canada, representing more than 44 million data points. The second issue is addressed by a proposed novel stochastic driving cycle construction method. The third issue limits the results to mainly Winnipeg and cities that have similar features, but the methodology can be used anywhere. The methodology develops the driving cycle using snippets extracted from recorded time-stamped speed of the vehicles from the collected database. The proposed Winnipeg Driving Cycle WPG01 characteristics are compared to eight existing standard driving cycles and are more able to represent aggressive driving, which is critical in PEV design. An attempt is made to isolate how many differences could be attributed to the sample size and the methodology. The proposed construction methodology is flexible to be optimized for any selection of driving parameters and thus can be a recommended approach to develop driving cycles for any drive train topology, including internal combustion engine vehicles, hybrid vehicles, plug-in hybrid, and battery electric vehicles. Characterization of vehicle parking durations and types of parking home, work, shopping, critical for duty cycles for PEV powertrains, are reported elsewhere. Here, the focus is on the mathematical approach to develop a drive cycle when a large database with high resolution of driving data is available.

Using the Pareto Set Pursuing Multiobjective Optimization Approach for Hybridization of a Plug-In Hybrid Electric Vehicle

A plug-in hybrid electric vehicle (PHEV) can improve fuel economy and emission reduction significantly compared to hybrid electric vehicles and conventional internal combustion engine (ICE) vehicles. Currently there lacks an efficient and effective approach to identify the optimal combination of the battery pack size, electric motor, and engine for PHEVs in the presence of multiple design objectives such as fuel economy, operating cost, and emission. This work proposes a design approach for optimal PHEV hybridization. Through integrating the Pareto set pursuing (PSP) multiobjective optimization algorithm and powertrain system analysis toolkit (PSAT) simulator on a Toyota Prius PHEV platform, 4480 possible combinations of design parameters (20 batteries, 14 motors, and 16 engines) were explored for PHEV20 and PHEV40 powertrain configurations. The proposed approach yielded the optimal solution in a small fraction of computational time, as compared to an exhaustive search. This confirms the efficiency and applicability of PSP to problems with discrete variables. In the design context we have found that battery, motor, and engine collectively define the optimal hybridization scheme, which also varies with the drive cycle and all electric range (AER). The proposed method and software platform could be applied to optimize other powertrain designs. [DOI: 10.1115/1.4007149]