Plants photosynthesise. Reptiles bask. The ancient Egyptians stored solar heat in the thick walls of their dwellings during the day to warm them at night. Archimedes is said to have used mirrors to burn ships at Syracuse in 213 BC.
The knowledge that electricity can come from sunlight has been with us since the 1830s, but it wasn’t until 1962 that a vintage 1912 Baker electric car, fitted with a solar array on its roof, showed the potential for solar transport by trundling along American roads at a sedate 30 km/h.
The first attention-grabbing demonstration of how sunlight could propel things, however, was not on the ground but in the air, when, in 1980, aerospace pioneer Paul MacCready’s solar-powered aircraft, Gossamer Penguin, took off. A year later, MacCready’s second, more sophisticated plane, SolarChallenger, crossed the English Channel.
Inspired by MacCready’s flights, Danish-born adventurer Hans Tholstrup and Australian racing car driver Larry Perkins undertook the world’s first transcontinental solar car journey, from Perth to Sydney, in the summer of 1981–82, laying the foundations for the biennial Darwin-to-Adelaide World Solar Challenge begun in 1987. Three years later, New Zealand joined the solar car fraternity when Hamilton solar pioneer Stewart Lister competed in the event with Solar Kiwi, and then in 1993 drove that vehicle’s successor from Cape Reinga to Bluff, albeit in a piecemeal fashion.
Yet, nearly two decades on, the attraction of acquiring energy seemingly for nothing as evidenced by the ubiquitous solar garden light, which casts its pale blue glow over even the most modest suburban driveways—has failed to translate into widespread enthusiasm for harvesting sunlight for travel. Solar cars on New Zealand roads have been conspicuous by their absence.
Perhaps, in these challenging economic times, cost and commercial opportunity could kickstart a wave of interest in solar/electric driving. In December last year, New Zealand’s latest sunlight-fuelled vehicle, SolarFern, drove the length of the country in 11 days and showed how frugal it can be to travel on the smell of an oil-free rag.
The Christchurch-based SolarFern team stems back to university days, when friends Rob Glassey, Barry Twigley and Brent Thompson were inspired by Hans Tholstrup’s trans-Australia World Solar Challenge. In 2005, Glassey became an official observer for this demanding engineering competition. He came home and convinced his friends that they should build a solar car too.
Glassey, an electrical engineer, masterminded the electronics. Twigley, a mechanical engineer specialising in mechatronics, took charge of the engineering and assisted with electronics. Thompson, a boat-builder by trade, brought his skills in the use of composite materials to bear on construction of the body. Thompson also runs a small business building propellers for light aircraft, and his aerodynamics experience proved invaluable in the creation of a “slippery” car shape.
In 2007, their car, SolarFern, competed in the World Solar Challenge. Due to a sponsorship glitch, they had to borrow the solar array from Lister’s 1993 Solar Kiwi. But, having returned this array to its owners after the race, the team needed a new power source for their 2011 journey from Bluff to Cape Reinga. One using the modern and enormously cheaper solar cells now available through advances in technology over the intervening years.
Cells were acquired as manufacturing seconds from Evergreen Solar in the United States. The problem with using ‘rejects’ is that functional cells and worthless cells look the same, but if you’re an informed purchaser, can find a knowledgeable and honest seller and don’t mind soldering—a pastime arguably as therapeutic as knitting—there are bargains to be had.
For an exacting end use such as powering a solar car, matching of cells is critical. Because the same current passes through every cell in a string, if one cell can’t keep up with the others, the voltage on that cell can reverse and it will start absorbing power rather than generating it.
SolarFern’s 652 cells weigh about 5 kg and cost around $1200. They were individually tested, graded and matched before being soldered together into an array stuck onto the car body and covered with a protective fibreglass and resin laminate (including a smattering of gold ‘wizard dust’ distributed by the Christchurch Wizard before the resin had properly cured).
The laminate, together with the wiring and other materials needed, brought the total cost of the photovoltaic array to around $2000—less than the annual petrol budget of most commuters. Unlike petrol, of course, solar cells can be used again and again.
The audacious journey was almost over before it started. Just days before the scheduled departure from Bluff, SolarFern’s motor controller overheated, destroying some critical components and putting paid to further travel until the problem could be diagnosed and fixed. Attempts to repair the circuit failed and the team had to wait for transistors to be sent from Singapore.
The team’s small campground cabin was transformed into an electronics workshop, co-opting the toaster as an electrical heat-sink. Three days later, the spare parts arrived, and (after a further dismaying failure) the motor controller was deemed ready for service—fifth time lucky.
On December 10, a day behind schedule, SolarFern departed from the Bluff signpost for Cape Reinga, taking the inland scenic route over the Lindis Pass. Brent Thompson took the controls for the first three-hour stretch. Sitting in SolarFern’s reclined driving seat is reasonably comfortable, though it’s fairly tight inside the canopy, with little headroom. Accelerator and brake controls are conventional, but there is no steering wheel. Instead, the vehicle uses side-stick steering and, though unorthodox, the method saves space.
When driving a solar car, the most pressing concerns are monitoring and managing the battery. Without a battery a solar car will run when the sun shines and stop when it doesn’t. Putting a battery in gives you the ability to manage your fuel input (sunshine) and store some when you’re stationary or have an excess, to help climb hills and travel further. Knowing the state-of-charge of a battery is fundamental to the strategy of driving a solar car.
SolarFern’s battery pack is made up of 528 lithium polymer cells, divided into 176 groups of three. Each cell is between 3.7 and 4.2 volts, depending on its state of charge, giving a total of around 130 volts for the whole pack. The battery can store up to 4 kWh, enough to power the car 350 km at 50 km/h on the flat, or 200 to 250 km over hills. Going faster—the car is capable of over 120 km/h—reduces the distance travelled dramatically.
Life in a solar car is more or less governed by available light, gradient, the angle of the sun—by small changes in small numbers. There is rarely power to spare and the performance of the vehicle can be drastically changed by something as ephemeral as a passing cloud.
On the first day, with clear skies, the car travelled 408 km to Omarama. The team had to watch motor temperature carefully during the climb up the Lindis Pass as the amperage drawn was 10 times the rating of the motor. Going down the other side was easier, described by the driver as “more like flying a glider than driving a car”.
Cloudy weather and head winds cut the second-day distance tally, and the team had to be content with reaching only Timaru, a mere 260 km.
On day three, with batteries less than half full and heavy traffic between Timaru and Ashburton, frequent recharging stops were needed. The team used a “minimum current target” strategy. Knowing energy use in that terrain to be about 14.5 Watt-hours per kilometre at 50 km/h, by estimating energy already stored plus sunlight energy expected during the day, they chose a target—in this case, Cheviot.
But in solar driving, targets can be frustratingly hit-or-miss. For SolarFern, the current needed to maintain 50 km/h on level grade is 5 amps. On the gradient out of Timaru, 5.5 amps was needed to maintain speed, but the solar array was delivering only 1.5 amps. It was with considerable driver relief that a brightening sky and a lessening gradient allowed break-even to be achieved at 56 km/h, using 4.4 amps with solar input to match—not bad in poor weather.
On the stretch from Ashburton to Christchurch, the team decided that, although reaching Cheviot was possible on stored battery power, the weather looked marginal—and the car is not well equipped for rain. With no advantage in pressing on while still having available battery space to soak up the day’s remaining energy, the solar array was propped up facing west on Rob Glassey’s Christchurch lawn.
The amount of energy captured by a solar cell is greatest when the cell surface is facing the sun directly. Deviation from this 90-degree angle of incidence reduces efficiency at an increasing rate as more sunlight is reflected and less absorbed. Early in solar car development, flat solar arrays were common, and many could be physically tilted to follow the angle of the sun during the day. These days, the vehicles are capable of higher speeds and aerodynamics plays a more defining role in design.
SolarFern’s array is divided into quarters, with each set of 150 to 180 cells arranged in a particular solar orientation—necessary to maximise energy-gathering potential. Each section of the array is controlled by its own maximum power point tracker (MPPT), an electronic device that juggles voltage and current output to allow the batteries to store the most power possible during fluctuating and uneven solar input.
Typically, the SolarFern array produces between 700 and 800 W, with over 1 kW possible under good direct lighting, and boosted by reflections from surrounding clouds. In dim, overcast conditions the power output can fall as low as 50 W—barely enough to power an incandescent lightbulb.
Individual cell efficiency is around 15 to 16 per cent, but because the cells are wired in series, a single damaged or obstructed cell can lower the total array efficiency to 10 per cent, resulting in a game of electrical cat-and-mouse to find and short-circuit the offender. When troubleshooting an underperforming array, the voltage of each group of cells is measured, and deviation from the expected values helps indicate problem areas. Cell surfaces are then scanned with an infra-red temperature sensor to detect any spots that are hotter than the rest. Typically, overheating cells are bypassed with a wire link soldered across them to prevent these cells absorbing energy from the rest, which also renders them useless.
SolarFern’s next challenge was the Hundalee hill between Cheviot and Kaikoura. Climbing the hill required changing the rear, driving wheel from the standard 50 cm-circumference wheel to a smaller 30 cm one. Since the electric motor is directly attached to the hub of the rear wheel, without any gearbox, the vehicle is essentially stuck in high gear all the time. The 30 cm wheel is the equivalent of shifting into low.
With the new wheel in place, the car sat at an odd-looking nose-high angle. Driver Barry Twigley—selected to be guinea pig for the steep ascent by virtue of being the lightest member of the team—quickly realised that although it was easy enough to see forward, the visual references for positioning the car width-wise on the road seemed “all wrong”. Bumps were transmitted through the rear suspension much more strongly, too. Twice the back wheel bounced completely off the ground, spinning to high speed, then making tread-shredding sounds on re-contacting terra firma. And speedometer calibration, running off the motor’s rotor position sensors, was thrown out by the change of wheel circumference ratio, requiring some challenging mental arithmetic.
Once at the summit, it was a clear run with the standard wheel configuration to Kaikoura, but with the Marlborough hills in store, and headwinds to boot, the day’s challenges were far from over. After a particularly long grind, Glassey, at the helm, reported a controller temperature of 70ºC, with the motor quietly roasting at 102º. Given that the motor’s rare-earth magnets can begin to demagnetise at around 125º, this was as hot as it was deemed safe to let it get. The car was stopped to cool and tilted towards the sun by propping one of the front wheels and the rear wheel on two large tool boxes. Another long, steep hill could be seen around the corner, so Twigley was called to the driver’s seat once more. Picton campground was finally reached in damp darkness, with hazard lights flashing.
Clouds and rain dogged the team in the North Island, each day a matter of waiting for windows of bright sky to charge depleted batteries. Visibility in rain was a problem, as SolarFern’s cockpit pod has no windscreen wipers.
Although cloud is a problem for solar cars, cold temperatures are not—as the temperature of a silicon solar cell goes down, performance goes up. (In early morning sun, when the night has been cold and there is a bit of dew on the panel, voltages are 15 to 20 per cent higher than in the middle of the day.)
Just before National Park, the solar array seemed noticeably lacking in performance. Shading one section at a time—akin to removing spark-plug leads to look for misfiring cylinders—revealed that the front right section wasn’t working at all. One MPPT had malfunctioned through water shorting out high-voltage terminals—perhaps the cause of a mysterious burning smell a couple of days earlier. Fortunately, a replacement was at hand, and with the array working properly the battery was soon nearly full.
Auckland came and went, and then the further hill challenges of Northland. As well as needing to deploy the smaller driving wheels—up the Brynderwyn Hills, for instance—drivers had to be careful on the downhill sections, too. The solar car, being aerodynamically ‘clean’, lacks the inherent speed-governing drag of a normal car travelling downhill and, if unchecked, will continue to accelerate under the force of gravity. Careful braking is necessary even on straight downhill stretches to stop the car attaining new land-speed records. (Regeneration using braking would have returned power to the batteries, but this was a complex design feature beyond the available time and resources for the trip.)
Team SolarFern’s 12th day on the road since Bluff dawned bright and hopeful. Thompson, up early, pointed the solar cells east along the Mangonui waterfront, angling the stands steeply to catch not just the low sun, but also the light reflecting from the big mirror of quiet water alongside.
Ten minutes north of Awanui and seemingly in the middle of nowhere, the team passed a tiny solar-powered cafe, to the obvious delight of its colourful owner.
Past Houhora, the cloud once again thickened ominously. By the time SolarFern made Waitiki Landing, now within about 30 km of the Cape, she was running on the ragged edge of minimal stored energy.
There is always uncertainty as to exactly how much energy remains in a battery when it is running at low levels. The Coulomb counting method for estimating battery state works by adding energy going in and subtracting energy coming out, dead-reckoning from a known battery capacity. As the SolarFern battery had not been full for some time, there was now significant uncertainty about whether the car had enough energy to reach its destination without a recharge.
But enough it was, and SolarFern reached the Cape Reinga carpark, the official end of the road, at 12:12 pm on Wednesday, December 21. For Barry Twigley, opening the canopy and standing, elated, on the seat, and for the others crowded around, this was the happy climax of all they’d endured on the journey, a memory never to be forgotten. Importantly, too, their efforts had provided a catalogue of considerations for future solar car design. Graeme Neho, kaumatua of Ngati Kuri, the iwi that administers New Zealand’s northern extremities, had granted permission for SolarFern to proceed to the lighthouse. So with Twigley back at the wheel, the solar car was driven down the walking track to the iconic landmark—as far north as it’s possible to go without a big splash—providing the final proof that sunlight alone is enough to travel the length of Aotearoa.