Integrated lithium ion and ultra capacitor bank for effective motor control

Project Title

Integrated lithium ion and ultra capacitor bank for effective motor control

Project Area of Specialization Electrical/Electronic EngineeringProject Summary

Aside from a few hydrogen fueled vehicles, all electric cars in the world are powered by Lithium-ion cells, combined to form a battery. Li-ion cells have appeared as the cell of choice for every automobile manufacturer, however each use a different cathode/anode chemistry and different electric motors. A problem that arises for each such vehicle is the degradation these cells go through over time which reduce the overall capacity of the pack as it is being used. A key reason for this degradation is the abuse the battery goes through, which is caused by high power demands during acceleration as well as during regenerative braking. These frequent charge discharge cycles cause side-reactions in the cell which causes the aforementioned cell degradation. If this isn't well constrained, the electric vehicle can lose a lot of usable capacity. Li-ion cells also have a reduced power capacity at low states of charge.

The proposed solution consists of a battery that is constructed such that Ultra-capacitors are coupled with the Lithium-ion pack. These are a new type of capacitors built from graphite carbon in the form of activated conductive carbon, carbon nanotubes. Generally, they have approximately 1000 times greater capacitance per volume compared to electrolytic capacitors. This has enabled them to be used in place on batteries in trucks to start with and the technology is being leveraged to help the pure Lithium pack in electric vehicles. Ultra-capacitors have a much larger voltage range, deep cycles, high power density and millions of charge/discharge cycles. The main advantage lithium ions have over them is the much higher energy density, which is why the two are used together.

The proposed project simulates the load the battery will be subjected to but adds an intelligent battery management system that uses the capacitors alongside the battery to best use both of them. High amounts of current in either direction is abusive to the battery, but a condition the ultra-capacitors are well suited for. The BMS will be constantly making decisions as to when power needs to flow to or from each source.

As Ultra-capacitors are quite expensive currently and scarce, the project will be of a small scale model for an EV and will be using electrolytic capacitors. Since the working principle is the same, theoretically, the model should work for Ultra-capacitors as well which are more practical for an EV due to improvements in their density.

Project Objectives

The goal of the project is to build a prototype of an effective hybrid energy storage system that is specifically designed for the application of electric vehicles. The battery pack, controller and power converter are designed with a load that is representative of an electric vehicle load, which is an electric motor. The load profile would also simulate a profile that is similar to the load profile a vehicle accelerating and deaccelerating for different driving conditions.

Ultra-capacitors have much lower energy densities than the already adopted Lithium-ion cells but they do have much higher power densities and more importantly, much higher charge/discharge cycles. Experiments demonstrate that using Ultra-capacitors in parallel with a lithium battery to power a pulsed load at a frequency of 1Hz 5A and 10% duty cycle, improves efficiency by 74% and power output increases 5 folds.

The results show that the proposed solution is likely to demonstrate great improvements for electric vehicle applications due to similar nature of the load, as discussed previously. Our solution is to add Ultra-capacitors in parallel with the main battery source but controlled by an intelligent system that is capable of deciding the power available by both the sources and switching between them to best optimize the power delivery to the load. In most cases, if the capacitor bank has power available, it will be used, the same for the case of regenerative braking and low state of charge for the bank. The bank has low energy capacity therefore its state of charge will quickly deplete during acceleration and vice versa for deceleration. This is important to remember as any battery's maximum power input and output is dependent on its state of charge

Estimating state of charge for each of these will require a lot of computation. It will need an accurate model for Li-ion cells, which includes open circuit voltage estimation, charge estimation as well as hysteresis estimation. This needs to be done individually for each cell in the pack. The model of the cell has a voltage source, 2 parallel resister-capacitor branches and a series resistor. Moreover, the capacitor bank also requires a model, which for the scope of this project will be a series capacitor and series resistance (ESR). The BMS will gather all relevant data for the available power decisions based on state of charge and decide which source needs to be used.

Project Implementation Method

The project can be broken down in 4 components and their implementation is listed:

1) Hybrid energy storage system: This will be a system of lithium ion cells and ultra-capacitors. All of them are connected in series. This gives higher voltage at the terminals while also reducing the number of current paths. This is beneficial as current sensors are much more inaccurate than voltage sensors and therefore gives better estimations for the BMS while the total energy and power remains constant. Another advantage of higher voltage levels is reduced current for the same power.

The lithium ion cells will be used to only supply power, to the load and the capacitors. They will not be charged via regenerative braking. The battery will charge the capacitors via a DC Boost converter module. The capacitors will be used as a power/energy buffer between the cells and the load and will supply the initial peak power demand and absorb energy from the motor, when simulating regenerative braking.

2) BMS: BMS will use voltage and sensors at every point in the energy storage system, taking voltage for at each node and current for each part. The sensors will be based on op-amps. All sensors are designed to output a voltage suitable for an Arduino’s analog input. An Arduino will process all the voltage values and compute the necessary data to be sent to the machine running the algorithm for state of charge estimation, power estimation and other aspects of the simulation. The BMS will be running a version of the extended Kalman filter for state estimation and compute the required control signal for the converter to control power flow in the circuit.

3) DC-DC converter: A bi-directional converter is being tested that will be able to control power being taken from the capacitors and the power taken from the motor during periods of regenerative braking.

While this converter is not the focus of the project, it is a key enabler to the entire project.

4) Load: The load being used to simulate is a MOSFET based resistive load. Changing the duty cycle of the switch will result in a simulated variable load and a separate power source to simulate regenerative braking. A motor based load would be unable to properly simulate different load conditions. Small motors available in the market are quite inefficient and would result in huge power losses, especially while simulating regenerative braking.

Benefits of the Project

A successful implementation of the project will result in a working prototype of the hybrid energy storage system (HESS) that will have significant advantages for all types of electric vehicles, from pure battery electric vehicles to mild hybrids. Currently, battery packs for hybrid vehicles need to be overdesigned to meet power requirements, which would not be the case for our proposed HESS. Ultra-capacitors may even fully replace batteries in hybrid vehicles. For pure battery EVs, such a system will improve performance of these vehicles by being able to supply the required power, which is not the suitable for lithium ion cells. Furthermore, regenerative braking will be more efficient as more energy would be able to be stored back and less wasted by brakes. Since many of the acceleration and deceleration cycles will be powered only through the capacitor bank, the system will also increase the life of the battery pack and reduce the overall cost of ownership of an electric vehicle. The ultra-capacitors have many more charge discharge cycles compared to lithium ion cells. A typical lithium cell has 3000-4000 cycles, an ultra-capacitor can have charge discharge cycles in the order of millions.

Technical Details of Final Deliverable

Lithium-ion battery pack: 6 18650 cells in series. Each cell has nominal voltage of 3.7V, can be charged up to 4.2V and discharged till a minimum of 3V. The rated capacity of each cell is 2600mAh. Maximum discharge current of the cell is 2.6A while maximum charge current is 2.0A.

The cells are manufactured by Samsung.

Capacitor Bank: 8 Capacitors in series. Each capacitor is rated for 350F and can be charged for up to 2.7V. Minimum voltage is 0V. Maximum charge/discharge current of the capacitors is 170A. The cells are manufactured by Maxwell Technologies.

BMS: voltages and currents are measured using differential op-amp configuration. For current measurement, series resistors are used. The gains for the voltage sensors is set to 1 using 1000 Ohm resistors while the gain of current sensors is set to 5. The sensors will be connected with the analog in ports of Arduino Mega. Due to limited number of analog connections, two will be used that will be connected via serial lines. Both of the Arduinos will act as slaves and a laptop computer will control them. This device will be running the algorithm, as well as implement the extended Kalman filter for state estimation. The device will also be responsible for presenting results.

Converter: The DC-DC will be using a bidirectional buck boost converter configuration with 2 MOSFETs. The duty cycle of the switch will govern the power flow and will be controlled by the computer, the Arduino Mega generating the required PWM signal. The inductor used would be iron powder core and in excess of 0.75mH to ensure converter operation in CCM. Maximum power output is 150W, which is the point at which the load is designed as well. The MOSFETs will be using driver ICs.

Load: Power resistors with switches in series will be used to simulate different load conditions and change the average power drawn from the load. This will be used to generate different load parameters.  A 3 series parallel branch with overlapping on states to ensure continuous power flow. A DC power supply will be used to simulate energy being taken form the vehicle and being absorbed by the capacitor bank.

Final Deliverable of the Project HW/SW integrated systemCore Industry Energy Other Industries Transportation Core Technology OthersOther TechnologiesSustainable Development Goals Industry, Innovation and Infrastructure, Sustainable Cities and CommunitiesRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	27600
Lithium ion cells	Equipment	6	200	1200
Ultra capacitors	Equipment	8	2200	17600
MOSFET	Equipment	10	30	300
MOSFET driver	Equipment	10	150	1500
Test Capacitors	Equipment	40	30	1200
Soldering equipment	Equipment	1	700	700
Arduino Mega	Equipment	2	1000	2000
LM324	Equipment	5	20	100
PCB	Equipment	3	1000	3000