Thursday, May 26, 2011

The Best EV-ER -- Electric Vehicle with Extended Range

Let us suppose that, through a combination of forces including governmental mileage or emissions mandates, market fuel prices, and technology maturation, it becomes de rigour for vehicle manufacturers to focus a good chunk of their energies on Electric Vehicles with Extended Range capacity (EV-ER). After all, while such vehicles are more complex than Battery Electric Vehicles (BEV), they sooth practical and imagined range-anxiety concerns by carrying some energy-dense fuel instead of a huge number of heavy and expensive batteries. Virtually all manufacturers are getting to understand BEVs -- regardless of the ever-changing flavors of battery chemistry, the key components of the EV system (motor, controller, charger, management system, electric peripherals, etc.) are pretty well-decided. However, the optimal ER device has not been established: in fact, it’s not even clear whether the ER device should only provide electricity (series design) or also provide motive power (parallel design). What are the manufacturers' options, and which is best?

Clearly, there are a variety of ER device options: 1. gasoline Internal Combustion Engine (ICE); 2. diesel ICE; 3. rotary ICE; 4. Homogeneous Charge Compression Ignition (HCCI) ICE; 5. two cycle ICE; 6. split intake and power stroke ICE; 7. micro-turbine; 8. sterling engine; 9. ethanol fuel cell; 10. hydrogen fuel cell.

And, there are just as many factors to consider in determining the pros and cons of each ER device: 1. does it define whether the vehicle must have a series design, as opposed to possibly allowing for parallel design; 2. manufacturing cost; 3. fuel economy; 4. fuel flexibility; 5. size; 6. weight; 7. noise, vibration, and harshness (NVH); 8. longevity/maintenance; 9. usability in the full spectrum of real world conditions; 10. implication for battery pack size.

The first factor, series versus parallel design, is critical, as it defines what kind of ER device you can use. For long highway driving, parallel architecture, which allows a spinning ER device to directly drive the wheels, would seem an advantage. Otherwise, you have to accept the series design’s somewhat Rube Goldberg-esque process of taking mechanical motion, converting it into electricity, shunting that electricity through a controller, storing it in the battery, pulling it back out of the battery, pushing it back through the controller, and running it through the motor to once again get mechanical motion -- and accept the loss of energy in each step. It has been estimated that these losses by themselves would add up to 10-15% energy loss. It might seem important to avoid losses like this, but is it?

First, engineers would point out that with a series design, in which the ER device is not directly tied to the wheels but instead simply makes electricity, it only needs to run at a single optimally-designed speed range and load. As a purpose-built generator, it may be efficient enough to make up for the 10-15% energy losses. Further, this design does away with the friction losses, complexity, expense, reliability, and weight issues of a mechanical transmission. It also vitiates the need to engineer for all those annoying on-throttle / off-throttle / part-throttle situations, and therefore remove complex fuel management, NVH, driveline lash, and a host of related “drivability” issues.

Is there a downside to series design? Of course. Series design requires a commitment to a large battery pack, with its attendant cost and weight issues, because while the parallel design can use the ER engine as well as the electric drive system to help move the vehicle when needed, the series design must have enough power in the battery pack and a large enough motor to move the vehicle by itself. And, as Toyota, GM, and others have implemented in variations of a blended parallel-hybrid design, the engine and traction motor are cleverly integrated into a planetary gearset so as to readily allow simultaneous motor and engine power transmission, smoothing out of NVH and other drivability issues (such as throttle response).

But wait, there's more to consider: interestingly, through-the-road parallel (i.e., one axle is driven by the motor, one axle is driven by the ER engine: Peugeot is the best (and only) example) does away with the need for a generator motor and offers 4-wheel drive capability to boot -- yet unfortunately, we’re back to requiring a transmission, and the coordination of the drives can be complex. Another alternative is employing the motor on the same drive shaft as the ER engine, which is a point in favor of parallel design as they can then both be reduced in size and used simultaneously when called upon for maximum acceleration: yet, the problem with this design is that if the ER only helps now and then, it can't heat up its catalytic converter and must then be run inefficiently to heat it up (electrically-heated catcons would still take a lot of energy).

Additional considerations: if you want more top speed, you need a more expensive motor in series designs due to the motor's (presumed) single-speed transmission -- a point in favor of parallel design. But, we anticipate improved batteries, which would slot nicely into a series design -- a point in its favor. A parallel design can provide a conventional sense of throttle-to-sensory response during highway driving -- a point in its favor. The modularity of a series enables the manufacturer to make both BEVs and EV-ERs on a single platform by simply removing the EF device and installing more batteries; or, switch ER devices as desired by the customer, or as best fits fuel availability, and as best fits other conditions found where that vehicle is to be sold -- all points for series design.

This last point may ultimately prove the convincer for manufacturers: one platform will enable architecture for a BEV, for improved batteries, for a choice of ER devises, and for modular development and selection of various equipment choices. For example, while GM has not officially announced that the Volt will be modified and sold as a BEV, it has taken steps in this direction.

Therefore, assuming a series design solution, we can look at all the remaining factors. (But given that the parallel v. series decision is still to be made by many manufacturers, it will be included in the analysis.)

GASOLINE
Parallel v. Series : either
Manufacturing Cost : medium (well known)
Fuel Efficiency : approx. 35%
Fuel Flexibility : gas or ethanol
Size : medium
Weight : medium
NVH : good
Longevity/Maintenance: complex but known
Real-World Usability : very good
Battery Size Needed : offers flexibility

DIESEL
Parallel v. Series : either
Manufacturing Cost : medium (well known)
Fuel Efficiency : approx. 40%
Fuel Flexibility : biodiesel
Size : medium
Weight : heavy
NVH : fair
Longevity/Maintenance: complex but known
Real-World Usability : very good
Battery Size Needed : offers flexibility

ROTARY
Parallel v. Series : better in series
Manufacturing Cost : medium
Fuel Efficiency : approx. 30%
Fuel Flexibility : gas or ethanol
Size : small
Weight : light
NVH : very good
Longevity/Maintenance: complex, less well-known
Real-World Usability : very good
Battery Size Needed : may offer flexibility

HCCI
Parallel v. Series : better in series
Manufacturing Cost : medium
Fuel Efficiency : approx. 40%
Fuel Flexibility : gas or ethanol
Size : medium
Weight : medium
NVH : good
Longevity/Maintenance: complex, less well-known
Real-World Usability : likely to be good
Battery Size Needed : may offer flexibility

TWO CYCLE
Parallel v. Series : better in series
Manufacturing Cost : medium
Fuel Efficiency : approx. 40%
Fuel Flexibility : gas or ethanol
Size : small
Weight : light
NVH : good
Longevity/Maintenance : complex, less well known
Real-World Usability : good
Battery Size Needed : may offer flexibility

SPLIT STROKE
Parallel v. Series : better in series
Manufacturing Cost : medium
Fuel Efficiency : approx. 40%
Fuel Flexibility : gas or ethanol
Size : medium
Weight : medium
NVH : good
Longevity/Maintenance: complex, less well-known
Real-World Usability : good
Battery Size Needed : may offer flexibility

MICRO-TURBINE
Parallel v. Series : series only
Manufacturing Cost : medium high
Fuel Efficiency : approx. 30%
Fuel Flexibility : excellent
Size : medium
Weight : light
NVH : low vibration, high noise
Longevity/Maintenance: very good
Real-World Usability : good
Battery Size Needed : large pack

STERLING
Parallel v. Series : series only
Manufacturing Cost : medium
Fuel Efficiency : approx. 40%
Fuel Flexibility : excellent
Size : large
Weight : medium
NVH : very good
Longevity/Maintenance: good
Real-World Usability : fair (slow to start)
Battery Size Needed : large pack

ETHANOL FUEL CELL
Parallel v. Series : series only
Manufacturing Cost : high
Fuel Efficiency : approx. 50%
Fuel Flexibility : ethanol only
Size : medium
Weight : light
NVH : excellent
Longevity/Maintenance: unknown, but should be excellent
Real-World Usability : unknown, but should be good
Battery Size Needed : large pack

HYDROGEN FUEL CELL
Parallel v. Series : series only
Manufacturing Cost : very high
Fuel Efficiency : approx. 50%
Fuel Flexibility : hydrogen only (query its source)
Size : medium
Weight : light
NVH : excellent
Longevity/Maintenance: unknown, but should be excellent
Real-World Usability : unknown, but should be good
Battery Size Needed : large pack

If a series design is chosen, and assuming there is reasonable exploration of ER devices, which device will become most common? The analysis demonstrates factors to consider: for instance, if the cost would come down, then direct ethanol fuel cells could be great -- but they obviously require pure ethanol, which is simply unavailable and won't be available without a serious policy commitment. Sterling engines could be great given their fuel flexibility, but they're large and difficult to manage due to their slow light-up: in many ways, it is a similar analysis with a micro-turbine. Rotary engines are small and light, but have never been broadly accepted and are not terrifically fuel efficient.

Yet another but: if these ER devices are used in series with large battery packs that will be primarily charged from the grid, does fuel efficiency matter that much? It may be that ownership will be a key issue: if the owner is able to charge daily, then the ER’s efficiency may not be a great concern as it may only seldom be used. (A study of early Volt adopters demonstrates that little gasoline is used.) On the other hand, if the owner winds up essentially running the vehicle on the ER device, then efficiency does matter. Finally, given the uncertainty of fuel sources in the future -- gas, diesel, bioethanol, biodiesel, butanol, natural gas, synthfuel, hydrogen -- are all combustible, and some of these ER devices can combust each of them. Clearly, trade-offs abound.

Taking into consideration a combination of factors, a purpose-designed series ER engine, such as the Lotus' "Omnivore" two-cycle, may seem a good way to go. It is relatively conventional in manufacture, highly efficient, not too weird for existing technicians (we stopped calling them "garage mechanics" when they started charging more than every other blue collar job), and small and light. But, it's not really omnivorous, as it does not profess to be able to burn diesel. Yet with the right manufacture, perhaps it could be designed to live up to its name, and then it could suffice for all possible fuels.

In the final analysis, the manufacturers are going to have make decisions that will affect the planet's travelers, and while everyone wants to look smarter than the other guy and drive in the fast lane, no one wants to drive down a dead-end road. Therefore, given the comfort level that existing engineers have with internal combustion, I think it is likely that the future will be BEVs and EV-ERs built on the same basic architecture to satisfy mass manufacturing, marketing, and modularity concerns, with the ER device comprised of one or another version of a small, light, efficient, fuel-flexible, and semi-conventional range extender. Vive la difference, as long as it's not too different.

No comments:

Post a Comment