Almost Great: GM EV1

First, it was built on a dedicated platform. The jury is still out on whether dedicated EV platforms or jack-of-all-trades ICE/hybrid/BEV platforms make the better business case, but I would argue that the purely technical argument is open-and-shut. To get the best out of EV technology, it’s hard to accept the inevitable compromises required to house large batteries, while at the same time providing homes for the pistons, fuel, air intakes and exhaust systems that explosion engines need. The fact that the most efficient EVs built to date – Tesla Model 3 and Lucid Air, take a bow – are both built on dedicated platforms illustrates my point.

And that key word ‘efficiency’ leads me neatly to the second box that the EV1 roundly ticked. Peter Savagian was one of the key engineers on the EV1. Initially responsible for the power electronics, later for the whole e-drivetain, Pete went on to have a stellar career within and outside of GM, but 35 years on he remembers the EV1 as ‘foundational for me as an engineer’. When he talks about it, you can see his eyes shining with enthusiasm for the product and the people. Pete told me that what remains with him is ‘the technical beauty of the concept – it was just a very efficient vehicle’. I can’t disagree. The EV1 was a lot more than a bunch of clever components – it was a master-class in system engineering, executed with a gimlet eye on whole-vehicle efficiency: my words, not his – Pete is far too modest a chap to make such claims.

The team of which Pete was a key part honed every vehicle metric that contributed to efficiency. The EV1 weighed in at between 1300 and 1400kg – featherweight for a BEV with the battery tech of the day – thanks to an innovative aluminium spaceframe, clad in composite-plastic panels, bonded in places to the spaceframe to share in the load-bearing work. Lots of clunky 1990s computer simulation and many hours in the wind tunnel resulted in an impressive drag coefficient of 0.19 – aided in part by the brave move of having a rear axle track over 200mm narrower than the front. Even more impressive, its creators paid a lot of attention to that second, much-neglected factor in drag reduction – frontal area. The EV1 had an astounding Cd*A figure of just 0.367 m². Xiaomi recently claimed a Cd figure of 0.195 for its SU7 a ‘best ever’ – Xiaomi has either forgotten or never heard of the EV1. And as usual, it neglected to publish the frontal area but trust me, it’s nowhere near as good as the EV1’s.

Pete has a great anecdote about the sheer depth of the aero work. One of his colleagues was a certain George Claypole, a brilliant aero-thermal engineer with a background in NASCAR – a slightly dishevelled but commanding figure. George and his team worked late shifts in the wind tunnel, working ungodly hours of the late night and early morning when the expensive facility was less in demand and they could maximise their time.

One night the team mocked up the latest design surfaces on the aero model – surfaces that had already been agreed by the Chief Vehicle Engineer and the design team. Much as they toiled through the night and early morning with blocks of foam, duct tape and cardboard, they simply could not get the new surfaces to work: precious hundredths of Cd had gone awol as compared to the earlier ‘draft’ surfaces. As dawn broke, and George nervously chewed his pencils, he took a remarkably brave decision. In those days the huge computer files that represented the 3D surfaces were not copied blithely on server after server – they occupied so much space in then-expensive memory that there were only two copies: one in the design studio, one sent to engineering. George ordered his team to simply delete the latest surfaces. Are you sure, boss? asked the team, conscious that this was one-way street. Just do it, came the reply. The team did as they were told, deleted the files and went home to grab a few hours’ sleep.

Some days later, George was called in by the project lead, to be told in no uncertain terms that he’d screwed up royally. He and his team has released the wrong – i.e. old – 3D surfaces to the body team, who had now launched the production tools. He’d made a helluva mistake – did he know that? ‘No I don’t,’ replied the unperturbable wind tunnel guru. ‘What I’ve done is saved you from making the biggest fuck-up of your career’. Mic drop, exit stage left. George was one of those rarest of birds – an engineer who cared more about the product and doing the right thing technically than following company orders or playing it safe. Pete cites him as one of his engineering heroes – and I think I can see why.

The GM engineers did not stop with breaking the rules in the wind tunnel. The EV1 sat on special 175/65 R14 tyres that were inflated to 50psi – thus helping both rolling-resistance and reducing frontal area. We modern EV engineers should hang our heads in shame as we roll over and meekly put huge tyres on our cars, just because they look cool.

One last example of the lengths the EV1 team went to eke every last yard of range out of each and every Watt-hour was its heating system. It was the first car to use a heat pump to scavenge the energy in the ambient air around the car in order to heat the cabin. Ten years later, a brilliant engineer called Jean-Marie L’Huillier came to me to propose a similar idea for the Renault ZOE. I thought he was nuts at first, but he convinced me, partly by educating me about the EV1, and I am very proud indeed that the ZOE became the second, and first mass-produced car, to use this beautifully simple and elegant way to make more of less.

So: dedicated platform, obsessional whole-vehicle focus on efficiency. Check, and check.

Third item – efficient and controllable motors. Now here is where things got really interesting on the EV1. It’s fairly intuitive that EVs need energy-dense batteries – that’s a bit of a no-brainer. But until the EV1, it was far from obvious just how to control the actual motors that turn the wheels. The earliest EVs – as well as milk-floats, golf cars and the like – used simple DC (direct current) motors with a variable resistance to vary the voltage applied to the motor’s windings: a horribly inefficient way of doing things. But in the 1980s, industrial engineers invented something called an AC motor variable-frequency drive.

This involved using a class of newly developed heavy-duty solid-state switches or transistors to switch an electrical supply on and off, very quickly. If they were switched in the right way, controlled by the cost-effective microcontrollers that were also coming onto the market in the 1980s, they could turn a DC power source, like a battery, into an AC or alternating current waveform. Not only that (and here’s the science bit), they could synthesise not just a fixed-frequency AC wave like the 50 or 60Hz one that will stand your hair on end if you stick your finger in the wall-plug, but a variable-frequency wave that could go from 0Hz to several kilohertz.

Now (still with me?) if you apply this variable-frequency AC waveform to the three phases of the windings of an EV’s motor stator, guess what happens? Well, the rotor – the bit attached to the wheels via a reduction gear and (preferably) a diff – will start to whirl at the same speed as the frequency of the applied waveform. If you attach the whole shooting match to a variable controller – let’s say, a pedal operated by your right big toe – you now have smooth, progressive control of your traction motor, all the way from 0rpm and 0mph, to stop-I-want-to-get-off Porsche Taycan Turbo S speeds.

This was the big technical breakthrough of the EV1 – it was the first EV to apply this motor-control technology to moving a car. Pete Savagian was the engineer responsible for these amazing power electronics – notably the use of a special and (back then) relatively new class of high-power transistor known as the IGBT or Insulated Gate Bipolar Transistor, and for the idea of liquid-cooling them to make sure they could handle the thermal stresses of switching very high currents very quickly. Every EV to the present day uses a version of this technology – generally and somewhat loosely dubbed an ‘inverter’ – to control the motor. For this innovation alone the EV1 deserves to go down as one of the greats.

Almost. Don’t forget the name of this series… Because we’ve come to the fourth and final of my EV tests – sadly, the one the EV1 did not ace. The battery.

Now, while few know and fewer care about the importance of the IGBT and the variable-frequency drive, most people know that one key driver for the resurgence of the EV was the availability of energy-dense, affordable batteries. EVs are all about cost-density of batteries – the famous $/kWh (dollar per kilowatt-hour) number that EV makers still obsess over. These days, this number has sunk as low as 60$/kWh* – a number that allows EVs to compete, albeit with great difficulty, with the incredible energy-cost-density of petrol or diesel. This level of cash-energy-density has been achieved thanks to the 1976 invention by a British scientist, one M. Stanley Whittingham, of the Lithium-ion battery – the little power-pack that fuels everything in your house from your iPhone to your laptop, those snazzy little earbuds or even the derailleurs on fancy bicycles.

But here’s the thing – GM did not choose this technology for the EV1. Instead, the first Gen 1 cars used ages-old lead-acid batteries, giving them a range of maybe 80 miles on a good day. The Gen 2 version, available from 1999, plumped for the Nickel Metal Hydride (NiMH) batteries that were also chosen by Toyota for the contemporary Prius – this chemistry near-on doubled the effective range to around 140 miles. Frustratingly, GM’s engineers – despite making so many of the right calls – choose the wrong tech. And it’s not as if it was not available. Li-ion cells had been around – as mentioned above – for more than 20 years. Nissan’s Altra EV, which was the first vehicle to use (Sony’s) Lithium-ion cells – was another close contemporary of the EV1, launched in even tinier numbers in 1998. Pete agrees that the EV1 missed out Li-ion, but also rightly points out that in any case even the nickel batteries were already too expensive – lithium cells would have been even more pricey at the time.

Still, it’s interesting to dream – if the EV1 has used Lithium-ion battery technology, who knows what the vehicle landscape might look like now? Maybe the General would have taken the lead in EVs. Maybe Elon would have stuck to writing fintech software, or just gone straight to Mars. Maybe all those EV1s would never have had an appointment with the crusher.

Speaking of crushers – what about the theories that Big Oil or Big Government killed the EV1? Peter, who was there, smiles and shakes his head. The real story is both more prosaic and simpler. The car simply lost money – the Gen 2 NiMH batteries required to give it its ‘just about enough’ range were still prohibitively expensive at the turn of the century. GM was just losing too much money on each car. If the General has been flush with cash at the time, maybe it would have been funded as a technology loss-leader for a few years, as a feeler to see how quickly battery costs might go down. But in the late 1990s GM was going through a tough patch – it was losing the minivan and truck sales battle in the US to the likes of Ford, and its shareholders were impatient. Not the time to indulge in a quirky little loss-making eco-car. So sorry to disappoint you – in this case there is no grassy knoll, no second shooter. Nobody killed the EV1 – it simply died, just another victim of the usual cruel car industry Darwinian rules of survival.

*cell-level costs, in China. Terms and conditions most certainly will apply.