(last updated: 10oct2011 19:49 GMT)
Update: Volkswagon have just announced a 313mpg vehicle, the XL1. High-efficiency Hybrid Electric Vehicles are now a proven concept, but have yet to reach mass-production.
There's a common mistake being made in Electric Vehicle design: the assumption that Batteries are a power source. Batteries are a power storage mechanism. The advantage of Electric Vehicles is therefore in the motor: high torque at low RPM, compact size, and efficiency ratings far in excess of a combustion engine: even an "inefficient" electric motor will achieve 80%, with figures of 95% and above being possible for designs such as the LRK-TorqueMax (such as the Hub Wheel Motor by Printed Motor Works Ltd, and the EV / Aircraft EMRAX Motor, by Enstroj)
Additionally, with mass-production of vehicles and vehicle design revolving around the combustion engine, parts that are designed for use in such vehicles are cheap, whilst components specifically designed for Electric Vehicle usage - even something as simple as a 2-speed gearbox - are costly.
So there are two key questions that need answering. The first is: how to re-use existing off-the-shelf vehicle components (drivetrain etc.) to save costs, and to make maintenance easier. The second is: how, again using as much off-the-shelf components, can a fuel source be turned efficiently into electricity, on-board the vehicle, within a reasonable size and weight? This article answers both these questions.
It's worthwhile setting reasonable limits on the cost and the weight of the vehicle, in order to make any EV - custom-build or conversion - actually feasible. Volvo's Environmental Concept Car has the right idea, but they developed a custom 2-speed gearbox and used a non-mass-produced Gas Turbine.
Electric Motors and associated controllers themselves are almost prohibitively expensive: a new 14kW 144 Volt Brushless DC Motor and its associated controller will set you back $2,500 excluding shipping. It therefore makes sense to keep the power down to under 15kW - possibly even only 10kW. With a good body design, decent tyres and keeping the weight down to under 500kg, 10kW can drive a lightweight vehicle at 60mph (putting that into context: most standard road cars lose about 10kW just through turbulence on the underside of the vehicle, which is one of the primary reasons why racing cars and EVs have aerodynamic undersides).
Additionally, by setting an approximate 500kg weight guideline, any vehicle which is around this limit does not require power steering or power assisted brakes. These components merely consume power and compensate for a cascading runaway design methodology so prevalent in combustion engine vehicles.
So by doing something as simple as arbitrarily setting a half ton weight, Mass Decompounding reduces the number of required components. Additionally, the cost of the remaining required components also goes down. All that's left is to find affordable components.
In evaluating the deployment of Electric Wheel Hub Motors, the problem is gradients (hills). To get a 500kg vehicle up a 1-in-4 gradient, assuming 100% efficiency and a 13in wheel radius, requires 400Nm (3600 lb.ft) of Torque. To expect a Direct Drive Electric Wheel Hub Motor to cope with 100Nm Torque and still be affordable is unreasonable. Kelly Controls have a High Torque version of their 4.5kW Wheel Hub Motor, which can do 100Nm if its ratings are exceeded by 50%, but that means that four expensive 6kW Motor Controllers are required, just to place the motors at risk from burn-out, and the top speed would only be about 40mph. Even if this worked out, mechanically, the costs are $600 each for the motors and $800 for the controllers: a total of almost £4,000 for motors and controllers. Therefore, sadly, despite being innovative and highly suited to electric bicycles, Hub Motors must be discarded as a viable option for an affordable Electric Vehicle.
The other option is geared motors of some description. Unfortunately, it's still not possible to meet reasonable speed expectations as well as coping with gradients, using fixed gearing. Even a single 10kW Brushless DC Motor, generating 25Nm of Torque, is a factor of 16 out (assuming 100% efficiency). Gearing such a motor down by 16:1 has the side-effect of dividing the maximum RPM of 3500 RPM by 16 as well (approx 220RPM), which would mean a top speed of only around 15mph on a 13in wheel radius. So, again: fixed gear ratios cannot be used, if a practical vehicle is required.
This leaves multi-speed gearboxes as the only viable option. CVT gearboxes, whilst viable on small vehicles such as snow-mobiles, are both high maintenance as well as inefficient (belt-drive). An innovative design which would suffice (such as the 2-speed automatic gearbox utilised by the Volvo Concept Car) is custom-built, and as such is outside of the average person's budget. Also, whilst tiptronic and multi-gear planetary gearboxes deployed in sports cars and racing cars are also ideal: again, they are outside of the average person's budget.
All factors and alternative options considered, it all basically points towards using an off-the-shelf mass-produced pre-existing combustion engine drivetrain, right from the clutch, through gearbox, to the differential and ultimately to the wheels. This doesn't sound particularly sexy or exciting, but it does actually mean that even a mechanic stands a chance of putting such a vehicle together, as well as being able to maintain it and find replacement parts.
All that is needed therefore is to pick a suitable drivetrain - one that was designed for a compact lightweight vehicle, with engine sizes of between 500 and 1300cc. Low-powered drivetrain technology has not really been seen on mass-produced "First World" vehicles for about 25 years, except until very recently. It is particularly ironic therefore to have to choose between a 70-year-old Citroen 2CV drivetrain or a 5-year-old Fiat Seicento drivetrain. Even the modern Fiat Seicento was designed to reflect the style and size of the original and famous Fiat 500. The advantage of picking, for example, a 2007 Fiat Cinquecento drivetrain over a Citroen 2CV one is that spare parts are more readily available.
A highly informative and insightful article has been written by a family-owned Racing Transmission Company in the U.S., here. In it, they describe how one City Municipality decided to convert a fleet of vehicles to Electric. They found that gear-changes were intolerable for the average driver, due to the higher torques available to Electric Motors. Two solutions present themselves:
The use of an Automatic Transmission is technically a superb solution. FB Performance were willing to innovate and to go through all the problems inherent with using Electric Motors: much higher torque, and an idling RPM of 0.
RPM-matching is a technique deployed by Truck Drivers, usually also involving double-declutching. This technique is very tricky to master, especially on down-shifts, as it requires all three pedals to be pushed down, for best results. So the ideal situation would be that the Electric Motor automatically alters its RPM when the Clutch Pedal is pushed, to match the Gearbox RPM, so that when the clutch is engaged, there is no "mashing".
To even consider attempting such a combined electronics and mechanical exercise, if a standard Combustion Engine was involved, would be a long-term complex Project. Challenges which would need to be overcome would include, but not be limited to:
It's probably best to stop there in order not to alarm the reader, because the electronic equivalent, using an off-the-shelf Electric Motor Controller and a standard Electric Motor, is a far simpler and far less risky proposition. The key difference is that Electric Motor Controllers are invariably already electronic computer-controlled (e.g. they are 3-phase Brushless DC Motor Controllers) and take a variable input voltage using a simple potentiometer connected to a foot pedal, or a thumb or wrist operated control.
The conclusion is, therefore, that whilst a straight drop-in replacement of an Electric Motor for a Combustion Engine is feasible, the majority of road users will find that it is challenging to adapt their driving style, but that many will persist and overcome that challenge, in return for a lower cost vehicle and the enormous environmental and financial benefits of reduced emissions and increased fuel economy. The alternative is the Automatic Transmission, with the family-owned company FB Performance being the only company in the world to have a proven Automatic Transmission successfully adapted for EV use. Lastly there is an R&D option to re-program a standard off-the-shelf Electronic Motor Controller to perform RPM-matching whilst the clutch is disengaged, during a manual gear change.
The key mistake made by EV designers is in assuming that it's appropriate to run solely and exclusively on batteries. General Motors "EV1" Car from 1999, as well as Volvo's "Hybrid" Concept Car from 2007 first recognised that there is another way. Unfortunately, it's still quite hard to get hold of compact efficient MicroTurbines. MTT Europe have a 15kW "Range Extender" unit which is being developed, but even this unit, thanks to simple thermodynamics, is only approximately 20% efficient. Wilson Turbopower are along the right lines, with their Ceramic Microturbine which has an astonishing efficiency figure to 50%, thanks to the large temperature differential.
However, whilst MTT are designing the 15kW unit to be compact and for use in cars, the Wilson unit is 300kW (despite an 11in blade diameter). Regardless, there is one overriding and final deciding factor that eliminates both these options: they are not mass-produced, mass-market products, and on that basis alone, sadly they have to be ruled out until they reach maturity.
Other power sources such as steam engines or steam turbines: again, are ruled out on the basis of either being not particularly efficient or not mass-produced. This really leaves only one sensible mass-produced power source left: Generators.
It seems crazy to consider putting in a petrol or a diesel generator into a car! Surely, that defeats the whole object of the exercise, especially if the Electric Vehicle is forced, for economic reasons, to use a drivetrain from an Internal Combustion Engine vehicle as well? The reasons become apparent only when looking closely at the figures, based around the possibilities which are available only when a vehicle's weight is under 500kg.
Take for example this SDMO 4000 generator. The rated power is 3.4kW, but look at the fuel usage: close to one gallon takes just under 5 hours to consume. So, even if that 3.4kW could be used to only power a vehicle at 30mph, that's an astonishing, eye-popping and unbelievable fuel economy figure of 150 miles per gallon!
All that is needed therefore is some basic maths which proves that 30mph can be sustained on a 3.4kW power budget. The vital factors involved are Rolling Resistance (tyres on the road) and Air Resistance (wind). According to How Stuff Works, Low Rolling Resistance Tyres are 1.5 to 2.5 times better than ordinary tyres. It makes sense, therefore to use those. Converting units out of the USA's preferred "Hectares per potato" figures, even at 55mph the power required is 0.7kW (700 watts) for a 500kg vehicle using Low Rolling Resistance Tyres. That's 20% of the power budget!
To put that into perspective, however: if the vehicle was 1500kg - the kind of weight of a standard family car - then that figure would be 2.5 times larger, which would be 50% of the power budget! The benefits of sticking to a 500kg weight become clear.
That just leaves air resistance as the major deciding factor. Using the formula shown on Wikipedia for Drag physics it is possible to work out the theoretical top speed of a vehicle which only has 3.4kW of theoretical power available. Assuming drivetrain losses of 10% reduces that to 3kW. Subtract a further 700 watts (appx) for Rolling Resistance, bearing in mind that this 700 watt figure is for 55mph. The numbers required for the formula are the frontal area (which can be assumed to be 1.0 sqm for a small 500kg vehicle); the Density of Air which is 1.184 kg/m/m/m at 25C, and a reasonable guess at an Aerodynamic Coefficient (the Wikipedia page says between 0.25 and 0.45 for a car: let us choose a coefficient of 0.3).
Putting all of these numbers in reverse into the formula, we need to take the cube root of: 2300 x 2 / 1.184 / 0.3, which works out to be 23.5 metres per second. That's 52.5mph! Even assuming a truly dreadful aerodynamic shape (a 0.45 coefficient), and a larger surface area (1.5 sqm), the figure still comes out at 40.1mph. Just to show what happens: let's assume only an 80% drivetrain efficiency, and also include the rolling resistance losses: a figure of 39.3mph is still achievable (which indicates how critical it is, with Electric Vehicles, to achieve high efficiency and good aerodynamics).
Earlier we estimated that it would be reasonable to assume a 30mph speed for the duration it took to empty the Diesel Generator's 4.3 Litre tank (4.8 hours). The somewhat incredible but dubious figure of 150mpg was therefore picked out of thin air. It almost beggars belief, therefore, for that to be shown to be an underestimate by nearly a factor of two, if the 52.5mph figure is achievable. Driving for a duration of 4.8 hours at 52.5 mph, consuming only 0.93 gallons, gives a whopping - and very confusing - 270 miles per gallon. Even with the truly dreadful aerodynamic shape and the larger surface area, the fuel economy figure comes out at 202 miles per gallon.
It's best to try to put this into perspective. Firstly: Low Rolling Resistance Tyres are required. Secondly: the road surface is assumed to be ideal (no pot holes, and not made of mud). Thirdly: there are assumed to be no hills (to assist or hinder). Fourth: there is assumed to be no wind (to assist or hinder). Fifth: a Drag Coefficient of 0.3 was assumed (worst case 0.45). Sixth: a front surface area for a small car was assumed, of 1.0 square metres (worst case 1.5 square metres). However, unfortunately, these are still perfectly reasonable assumptions to make, so it is hard to see why the fuel economy figures can even be legitimate or correct, against a background of standard vehicle design and "modern" fuel economy figures.
The major parts are:
The most expensive of these is, surprisingly, the Generator! 5kW Diesel Generators retail for around £1500. The Motor Controller is second, at around £1000. The actual electric motor itself is third, at about £800 (depending on power), the Battery Charger is next at around £500, and the donor vehicle could likely be picked up for under £400, or as an MOT failure for under £250. Last comes a small battery pack, which, depending on size and the voltage required, would be around £380 for an EV-suitable (and, importantly, new and therefore trustworthy) battery pack.
(Update: there exists a very good donor vehicle already, called the Aixam City Car. It is already Category L7E and has been in production for over 10 years).
If, however, a new Kit Car is picked as the basis for the vehicle, then the Kit Car comes in as the top-priced item. The primary justification for picking a Kit Car (such as the Joyrider from Blitzword) is that it is light-weight, exciting, and, after passing an SVA test for road-legal purposes, qualifies as a "New Car", and thus does not require an MOT for three years. Even for the Kit Car, however, the drivetrain is selected from a range of suitable donor cars, where parts are readily available.
So the electrics comes in at around £4,000, whilst the actual vehicle that all sits in could be between £400 for a used vehicle, or up to £7,000 for a new one (as a Kit Car). A DIY vehicle could therefore be had for about £4,500 and a road-legal ready-to-drive one made for under £12,000.
The above system is known as a Series Hybrid powertrain. Particularly compelling is the story of the General Motors EV1 which, with a weight of 1400kg, a 137HP electric motor and a 40kW Williams International Gas Turbine Unit, could easily achieve between 72 and 120mpg.
The benefits of Series Hybrid designs are numerous.
Thanks to many examples, the GM EV1 and the Volvo ECC being just two, the feasibility and efficiency of Series Hybrid drivetrains has never been in doubt. What hasn't really been made clear is that it's perfectly possible for the average mechanic to get together with an average electronics engineer to put together a Series Hybrid vehicle using standard off-the-shelf mass-produced components, without having to spend an absolute fortune in parts or maintenance.