A Solar-Powered Car?

If you like the sun, and you like cars, then I’m guessing you’d love to have a solar-powered car, right? This trick works well for chocolate and peanut butter, but not so well for garlic bread and strawberries. So how compatible are cars with solar energy? Do we relish the combination or spit it out? Let’s throw the two together, mix with math, and see what happens.

What Are Our Options?

Short of some solar-to-liquid-fuel breakthrough—which I dearly hope can be realized, and described near the end of a recent post—we’re talking electric cars here. This is great, since electric drive trains can be marvelously efficient (ballpark 85–90%), and immediately permit the clever scheme of regenerative braking.

Obviously there is a battery involved as a power broker, and this battery can be charged (at perhaps 90% efficiency) via:

  • on-board internal combustion engine fueled by gasoline or equivalent;
  • utility electricity;
  • a fixed solar installation;
  • on-board solar panels.

Only the final two options constitute what I am calling a solar-powered car, ignoring the caveat that hydro, wind, and even fossil fuels are ultimately forms of solar energy. The last item on the list is the dream situation: no reliance on external factors other than weather. This suits the independent American spirit nicely. And clearly it’s possible because there is an annual race across the Australian desert for 100% on-board solar powered cars. Do such successful demonstrations today mean that widespread use of solar cars is just around the corner?

Full Speed Ahead!

First, let’s examine the requirements. For “acceptable” travel at freeway speeds (30 m/s, or 67 m.p.h.), and the ability to seat four people comfortably, we would have a very tough job getting a frontal area smaller than 2 m² and a drag coefficient smaller than cD = 0.2—yielding a “drag area” of 0.4 m². Even a bicyclist tends to have a larger drag area than this! Using the sort of math developed in the post on limits to gasoline fuel economy, we find that our car will experience a drag force of Fdrag = ½ρcDAv² ≈ 250 Newtons (about 55 lbs).

Work is force times distance, so to push the car 30 meters down the road each second will require about 7,500 J of energy (see the page on energy relations for units definitions and relationships). Since this is the amount of energy needed each second, we can immediately call this 7,500 Watts—which works out to about ten horsepower. I have not yet included rolling resistance, which is about 0.01 times the weight of the car. For a super-light loaded mass of 600 kg (6000 N), rolling resistance adds a 60 N constant force, requiring an additional 1800 W for a total of about 9 kW.

What can solar panels deliver? Let’s say you can score some space-quality 30% efficient panels (i.e., twice as efficient as typical panels on the market). In full, overhead sun, you may get 1,000 W/m² of solar flux, or a converted 300 W for each square meter of panel. We would then need 30 square meters of panel. Bad news: the top of a normal car has well less than 10 square meters available. I measured the upward facing area of a sedan (excluding windows, of course) and got about 3 m². A truck with a camper shell gave me 5 m².

If we can manage to get 2 kW of instantaneous power, this would allow the car in our example to reach a cruising speed on the flats of about 16 m/s (35 m.p.h.). In a climb, the car could lift itself up a grade at only one vertical meter every three seconds (6000 J to lift the car one meter, 2000 J/s of power available). This means a 5% grade would slow the car to 6.7 m/s, or 15 miles per hour—in full sun. Naturally, batteries will come in handy for smoothing out such variations: charging on the downhill and discharging on the uphill, for an average speed in the ballpark of 30 m.p.h.

Image from toyota.com.

So this dream of a family being comfortably hurtled down the road by real-time sun will not come to pass. (Note: some Prius models offered a solar roof option, but this just drove a fan for keeping the car cooler while parked—maybe simply offsetting the extra heat from having a dark panel on the roof!) But what of these races in Australia? We have real-live demonstrations.

The Dream Realized

Tokai Challenger. Look at that speed!

In recent years, the Tokai Challenger, from Tokai University in Japan, has been a top performer at the World Solar Challenge. They use a 1.8 kW array of 30% efficient panels (hey—my guess was right on!), implying 6 square meters of panel. The weight of the car plus driver is a mere 240 kg. As with most cars in the competition, the thing looks like a thin, worn-down bar of soap with a bubble for the driver’s head: both the drag coefficient (a trout-like 0.11) and the frontal area (I’m guessing about 1 m², but probably less) are trimmed to the most absurd imaginable limits. From these numbers, I compute a freeway-speed aerodynamic drag of about 60 Newtons and a rolling resistance of about 25 N, for a total of 85 N: about 35% of what we computed for a “comfortable” car. Solving for the speed at which the combination of air drag plus rolling resistance requires 1.8 kW of power input, I get 26 m/s, or 94 km/h, or 58 m.p.h., which is very close to the reported speed.

Bring on the Batteries: Just Add Sun

We have seen that a practical car operating strictly under its own on-board power turns in a disappointing performance. But if we could use a large battery bank, we could store energy received when the car is not in use, or from externally-delivered solar power. Even the Australian solar racers are allowed 5 kWh of storage on board. Let’s beef this up for driving in normal conditions. Using today’s production models as examples, the Volt, Leaf, and Tesla carry batteries rated at 16, 24, and 53 kWh, respectively.

Let’s say we want a photovoltaic (PV) installation—either on the car or at home—to provide all the juice, with the requirement that one day is enough to fill the “tank.” A typical location in the continental U.S. receives an average of 5 full-sun hours per day. This means that factoring in day/night, angle of the sun, season, and weather, a typical panel will gather as much energy in a day as it would have if the high-noon sun persisted for five hours. To charge the Volt, then, would require an array capable of cranking out 3 kW of peak power. The Tesla would require a 10 kW array to provide a daily charge. The PV areas required vastly exceed what is available on the car itself (need 10 m² even for the 3 kW system at a bank-breaking 30% efficiency; twice this area for affordable panels).

But this is not the best way to look at it. Most people care about how far they can travel each day. A typical electric car requires about 30 kWh per 100 miles driven. So if your daily march requires 30 miles of round-trip range, this takes about 10 kWh and will need a 2 kW PV system to provide the daily juice. You might be able to squeeze this onto the car roof.

How do the economics work out? Keeping up this 30 mile per day pattern, day after day, would require an annual gasoline cost of about $1000 (if the car gets about 40 MPG). Installed cost of PV is coming in around $4 per peak Watt lately, so the 2 kW system will cost $8000. Thus you offset (today’s) gas prices in 8 years. This math applies to the standard 15% efficient panels, which precludes a car-top solution. For this reason, I will primarily focus on stationary PV from here on.

Practicalities: Stand-Alone or Grid-Tie?

Ah—the practicalities. Where dreams get messy. For the purist, a totally solar car is not going to be so easy. The sun does not adhere to our rigid schedule, and we often have our car away from home during the prime-charging hours anyway. So to stay truly solar, we would need significant home storage to buffer against weather and charge-schedule mismatch.

The idea is that you could roll home at the end of the day, plug up your car, and transfer stored energy from the stationary battery bank to your car’s battery bank. You’d want to have several days of reliable juice, so we’re talking a battery bank of 30–50 kWh. At $100 per kWh for lead-acid, this adds something like $4000 to the cost of your system. But the batteries don’t last forever. Depending on how hard the batteries are cycled, they might last 3–5 years. A bigger bank has shallower cycles, and will therefore tolerate more of these and last longer, but for higher up-front cost.

The net effect is that the stationary battery bank will cost about $1000 per year, which is exactly what we had for the gasoline cost in the first place. However, I am often annoyed by economic arguments. More important to me is the fact that you can do it. Double the gas prices and we have our 8-year payback again, anyway. Purely economic decisions tend to be myopic, focused on the conditions of today (and with some reverence to trends of the past). But fundamental phase transitions like peak oil are seldom considered: we will need alternative choices—even if they are more expensive than the cheap options we enjoy today.

The other route to a solar car—much more widespread—is a grid-tied PV system. In this case, your night-time charging comes from traditional production inputs (large regional variations in mix of coal, gas, nuclear, and hydro), while your daytime PV production helps power other people’s air conditioners and other daytime electricity uses. Dedicating 2 kW of panel to your transportation needs therefore offsets the net demand on inputs (fossil fuel, in many cases), effectively acting to flatten demand variability. This is a good trend, as it employs otherwise underutilized resources at night, and provides (in aggregate) peak load relief so that perhaps another fossil fuel plant is not needed to satisfy peak demand. Here, the individual does not have to pay for a stationary battery bank. The grid acts as a battery, which will work well enough as long as the solar input fraction remains small.

As reassuring as it is that we’re dealing with a possible—if expensive—transportation option, I must disclose one additional gotcha that makes for a slightly less rosy picture.  Compared to a grid-tied PV system, a standalone system must build in extra overhead so that the batteries may be fully charged and conditioned on a regular basis.  As the batteries approach full charge, they require less current and therefore often throw away potential solar energy.  Combining this with charging efficiency (both in the electronics and in the battery), it is not unusual to need twice the PV outlay to get the same net delivered energy as one would have in a grid-tied system.  Then again, if we went full-scale grid-tied, we would need storage solutions that would again incur efficiency hits and require a greater build-up to compensate.

A Niche for Solar Transport

UCSD golf cart with PV roof.

There is a niche in which a vehicle with a PV roof could be self-satisfied. Golf carts that can get up to 25 m.p.h. (40 km/h) can be useful for neighborhood errands, or for transport within a small community. They are lightweight and slow, so they can get by with something like 15 kWh per 100 miles. Because travel distances are presumably small, we can probably keep within 10 miles per day, requiring 1.5 kWh of input per day. The battery is usually something like 5 kWh, so can store three days’ worth right in the cart. At an average of five full-sun hours per day, we need 300 W of generating capacity, which we can achieve with 2 square meters of 15% efficient PV panel. Hey! This could work: self-contained, self-powered transport. Plug it in only when weather conspires against you. And unlike unicorns, I’ve seen one of these beasts tooling around the UCSD campus!

Digression: Cars as the National Battery?

What if we eventually converted our fleet of petroleum-powered cars to electric cars with a substantial renewable infrastructure behind it. Would the cars themselves provide the storage we need to balance the system? For the U.S., let’s take 200 million cars, each able to store 30 kWh of energy. In the extreme, this provides 6 billion kWh of storage, which is about 50 times smaller than the full-scale battery that I have argued we would want to allow a complete renewable energy scheme. And this assumes that the cars have no demands of their own: that they obediently stay in place during times of need. In truth, cars will operate on a much more rigorous daily schedule (needing energy to commute, for instance) than what Mother Nature will throw at our solar/wind installations.

We should take what we can get, but using cars as a national battery does not get us very far. This doesn’t mean that in-car storage wouldn’t provide some essential service, though. Even without trying to double-task our electric cars (i.e., never demanding that they feed back to the electricity grid), such a fleet would still relieve oil demand, encourage renewable electricity production, and act as load balancer by preferentially slurping electricity at night.

I Want a Solar-Powered Car

I also want a land speeder from Star Wars, a light saber while we’re at it, and a jet pack. And a pony. But unlike many of these desires, a solar powered car can be a practical reality. I could go out tomorrow and buy a Volt or a Leaf and charge it with my home-built off-grid PV system (although I would first need to beef it up a bit to cover our modest weekly transportation needs). Alternatively, I could park a solar-charged golf cart in the sun—or charge an electric-assist bicycle with a small PV system, for that matter—to get around my neighborhood. Slightly less satisfying, I could install a grid-tied PV system with enough yearly production to offset my car’s electricity take. The point is, I could make stops at the gas station a thing of the past (or at least rare, in the case of a plug-in hybrid).

So solar powered cars fall solidly on the reality side of the reality-fantasy continuum. That said, pure solar transport (on board generation) will suffer serious limitations. More reliable transport comes with nuances that may be irritating to the purist. You can apply a bumper sticker that says SOLAR POWERED CAR, but in most cases, you will need to put an asterisk at the end with a lengthy footnote to explain exactly how you have realized that goal.

Hits: 10401

49 thoughts on “A Solar-Powered Car?

  1. Tom: It has occurred to me that the “cars as a national battery” would be much more expensive than just having the battery in your garage attached to the grid: the battery is much cheaper than the car. An arrangement that the power company would pay the battery-on-grid homeowner could amortize a battery much sooner than amortizing a battery powered car. A revenue-generating strategy for the homeowner, just like having the power company pay the homeowner for excess power generation. As I have commented to an earlier post on Do The Math, the current arrangement of excess solar power subtracted from your electrical bill but no opportunity to get a check from PGE makes roof top solar far less attractive.

  2. As a solar panel and Nissan Leaf owner I have a few comments. The Leaf does between 3 and 4 miles per KWh in real world use. The heater makes about 5% difference amd the AC makes about 10%. 3 miles/KWh for high speed and hilly driving, 4 m/KWh for urban speeds on the flat.

    Our grid tied solar installation is on PG&E E6 time of use plan. This varies from 8.5c/KWh off peak to 28c/KWH (approx – I forget the exact no) in the afternoons. This means that a few hours of sunny afternoon pays for many more hours of overnight charging. The costs are also tiered, so higher than baseline total electricity use increases the cost per unit. That makes it hard to get exact calculations, but most of the savings from a solar install come from moving the whole house down to the base usage level.

    We also switched appliances from gas to electricity and put in a heat pump and added AC to the house. Some of the return on that investment is an increase in the value of the house, not just a reduction in fuel costs.

    The net result is a 10KW solar array that runs the house and the Leaf and very low total fuel costs, and an investment that makes the house more valuable, just as a remodeled kitchen would.

  3. What are your thoughts on “Solar Roadways”?

    It’s obviously not feasible in any “normal” timeframe, but if at all possible it sounds like something that could help a great deal with at least some of the issues.

  4. Another excellent example and documentation of the problem the world faces as the virtually free energy density which was available to design and build modern civilization, the civilization as we in the OECD assume is ‘normal,’ fades into history.

    Fast, comfortable, personal overland transport is a construct of the oil era, nothing more or less.

    Overland commerce – the transport of goods at mass scale – required coal and rail before becoming part of the paradigm we accept as ‘normal.’ I’ve calculated the power and energy needs of a typical loaded semi-tractor trailer truck and solar power as described above isn’t realistic – or better said, trucking isn’t realistic.

    Given the deliberate destruction of 90%+ of the rail and local warehousing infrastructure in the US over the past 100 years, a workable transition from 41kW per gallon atmospheric-pressure liquid fuel powered economics to a system using anything less energy dense won’t be an option for the “consumer” masses that we’ve designed and evolved into over the past 150 years.

    That said, I’m another reader who generates over 30kWh/day locally. We do what we can to lighten the load.

    • The key to mass transportation is finding a sustainable liquid energy storage.
      Before the electro/hybrid car hype (not that it is all bad but the battery is a big problem) here in Germany the car companies worked hard on establishing Hydrogen as future fuel. As far as I know they still work on that but it has become quiet.
      The idea was that countries with high solar(wind/water would work too) power produce H2 and export it.
      One advantage: Airplanes can fly with H2 with range and payload in todays magnitude.
      The big question is if it would be possible to produce enough H2 in a sustainable way.

      • No airplanes can not fly on H2 today anywhere with the same range and payload like they do on jet fuel. The volumetric energy density of H2 is far to low to allow this around 1/3-1/4 that of jet fuel. It is a physics issue not a wishful thinking issue and physics trumps economics and wishful thinking every time and in the case of H2 will continue to do so until people get the message that is will not work for energy use.

        • Volumetric energy density may or may not be a limiting factor for aviation fuel, depending on aircraft design, but specific energy density always is. Hydrogen tops the list of course for specific energy density of chemical fuels, hence its use in rocket propulsion where large vertical tanks add plenty of volume without adding much drag.

    • Don’t write it off too hastily: comparing annual insolation for 239 U.S. cities over 30 years leads to the startling conclusion that the best place in the continental U.S. (Mojave Desert) outperforms the worst place (Olympic Peninsula) by only a factor of two! Granted, there is a large seasonal variation. But still. I was surprised. Are you surprised? Not the order of magnitude that one might expect (and I lived three years in Seattle, so I get it).

      You can find the data here.

      • Yes, that is quite surprising! Thanks for the info.

        FWIW, I wouldn’t write off solar electricity for residential and industrial uses even here, and I myself had a pretty worthwhile (for 3/4 of the year) solar HW panel on the house I used to own here.

        But it strikes me as self-defeating to propose squandering whatever capacity we can create trying to perpetuate cars-first transportation. But that’s kind of your point, isn’t it?

  5. It’s too bad you ignored acceleration when determining the power requirements. Super cars do 0-60mph in under 5 seconds. Without exact numbers, I’ll assume that minivans can do it in under 20 seconds. Let’s say that we’re willing to do 0-30m/s in 30 seconds to make the math easy. Assuming constant acceleration (HA!), we need to accelerate at 1m/s/s. F=ma, so the force required to get us up to speed is 600kg*1m/s/s=600N. P=Fd/t and d=t(v_max + v_min)/2 under constant acceleration, so the average power required to get us up to highway speeds, ignoring drag, is F*v_max/2 = 600N * 30m/s / 2 = 9kW.

    I alluded to the fact that constant acceleration in cars of any type is unrealistic. Lucky for us, using a constant 9kW of power solely for acceleration will still get us to 30m/s in 30s. We just won’t have constant acceleration. This is probably the lower limit of peak performance most people will accept. Even if we shift some of the power use to lower speeds during acceleration, we’re still limiting our high speed performance to a large degree for the sake of making it to 30m/s in a reasonable time frame. Doubling our power budget allows us to use a constant 9kW for acceleration, but that’s not very economical. The ideal number is close to 13kW.

    • I started a paragraph on acceleration based only on solar input, but realized that real cars wouldn’t work that way: they all have a battery and this allows bursts of power to handle acceleration. These lightweight cars might in fact feel rather zippy. But it’s certainly true that under solar alone, the acceleration time would seem glacially slow.

  6. Thank you for another nicely realistic assessment of one of our energy options.
    It is good to see someone actually using maths to figure this stuff out. It may be
    ‘back of the envelope’ type calculation, but that’s better than much of the pure
    handwaving argument I have read elsewhere.

    I have done the calculations to convert my Prius to being a plug in hybrid and
    connecting it to the 2kW PV array on the roof. It turns out to be uneconomical
    to do at the moment. One of the main issues is that the car needs to recharge
    overnight (when it is parked), which means using grid power or having to install lots of storage. The other main issue is the sheer cost of conversion. I saw a plug in Prius in a Paris Toyota showroom earlier this year, I may have to wait a bit.

  7. One potential way of would be to split the location of the solar panels – for example, many cars are parked out in a sunny carpark all during the day. So, you charge the car from the (presumably renewably sourced) grid or your home roof PVs at night for the commute to work, then the PVs on the car roof provide the juice for the drive home.

    I would certainly like to see more use of the solar golf buggies. And, hey, I commute by bicycle at somewhat less than 25mph, so that would actually be an improvement in speed on that.

    • Or, we could switch to much more efficient mass transit, especially in large cities. It would be win-win since not only would we decrease energy consumption but also we would stop cluttering small areas with more and more cars. It’s a nonsense to move 2-ton vehicle around just to move 1 or 2 people.

      • Though helpful for congestion, not all mass transit saves energy per passenger mile over the automobile.

        Energy used (BTUs) per passenger per mile:
        Cars: 3538 (decreasing at 0.8% per year)
        Personal Trucks: 3663
        Motorcycles: 2460
        Buses: 4242 (*increasing* at 1.4% per year)
        Rail, commuter: 2812

        Table 2.12

        So by all means jump aboard that *existing* half full bus instead of taking the car to save energy, just don’t buy a *another* bus for that purpose alone.

        • If the bus goes half empty and every year more and more people use cars instead of the bus, it’s no wonder that per passenger energy cost is high and increasing. Ultimately just the driver remains in the bus and “per passenger” BTU will be infinity.
          What I’ve proposed in another comment is a city-wide system that optimizes mass transit’s load. If it KNOWS how many people are going to be on that bus stop, it can send there a smaller bus if a big one is not needed.

  8. I don’t understand why you apply the “batteries don’t last forever” argument to the “buffer” stationary battery bank but not to the batteries in the car itself. As lithium batteries are much more expensive than lead-acid batteries, it is of even greater importance. With current lithium battery technologies, I wouldn’t be surprised that it is actually the dominant factor in cost estimates. The same issue also makes “V2G” schemes uneconomical.
    It is not an easy math exercise, as batteries who lost too much capacity to be useful in a car could still be used in other applications, such as grid storage, and fetch therefore a significant price, and completely worn-out lithium batteries should be quite valuable for recycling purpose.

    • It appears large automotive battery packs of LiFEPO batteries should obtain more than 3000 cycles before falling to 80-90% of capacity. If the pack size is 100 miles (~25kWh), then the total mileage on the battery is nearly 300,000 miles, likely longer than that of the vehicle, and then the battery owner may indeed decide to sell some cycles back to the grid given the option.

  9. Tom,
    I appreciate that many of your calculations show the absurdity of attempting to continue business as usual. Conversely, substantial changes in culture would have a profound effect on these calculations. For example, although I have had difficulty finding data on average passengers per car in the U.S., just watching traffic gives the impression that the large majority of vehicles are still driven with a single passenger. The assumption that a commute vehicle “needs” to accommodate 4 passengers comes into question. This is a cultural choice, and cultures can evolve. An alternative transportation choice that would require cultural change but have a huge payoff could be a shift to aerodynamic full electric motorcycles. http://www.brammo.com/home/ http://www.velomobiel.nl/allert/Recumbent%20motorbike.htm
    I wonder what your solar requirements would be like if say half the motor fleet where to move to this form of travel. As a side benefit, the total resource requirements would be substantially less than continuing to make vehicles that are 4 times larger than most people actually need.

    • Well, if we’re talking about that level of change, imagine what it would be like with real public transit systems, ones in which cars are as free and unrestricted as they are today but 80% of the public chooses to use transit 80% of the time because it’s the superior choice.

      I envision a system that starts with small-route neighborhood circulators (we have a few here in Tempe) that feed into the major arterial streets. In the Phoenix metro area, we’ve got a one-mile grid of arterials, perfect for a bus fleet. Then add high-speed commuter rail down the middle of the existing freeway system, and run everything at a minimum of ten-minute intervals for the most out-of-the-way off-peak routes. Just to handle capacity, you’re probably looking at one- or two-minute intervals for busy routes during rush hour.

      For riders, you don’t have to worry about schedules or even routes. Something will be along in just a few minutes. Ride it in the general direction you’re heading, get off when it starts to go the worng direction, and in a couple minutes another something-or-other will be along to get you back on track. And since we’d be replacing dozens of cars with a single bus, congestion would vanish, meaning everybody would get where they’re going in far less time than today. Factor in commuter rail which could easily top 100 MPH where you’re currently lucky to do 40 during rush hour and there’s no comparison. Plus, you can read or do work or whatever during your commute and not have to worry about traffic.

      Would such a system be expensive? Well, compared to what? The next time you’re in a parking lot, count the number of cars, multiply by $25,000 for the capital expenditure, add say 20% for financing over the lifetime of the purchase, add in $1000 per year for fuel, and another $1000 for insurance and fees…and it quickly becomes apparent that what we currently spend on transportation could buy us a solid-gold public transit system. I’ve seen single mall parking lots worth more than the entire Valley Metro system.

      The car certainly has its place in American society, and I have no desire to get rid of it. But we’re insane to build our transportation system around single-occupancy multi-passenger passenger-operated non-autonomous vehicles. It’s about the most inconvenient, slowest, least inefficient, most polluting, and most expensive system one could design.



  10. This is a good analysis but I find that in analyzing these alternatives we often tend to get into an all-or-nothing-mindset, and assume that technology and prices won’t progress as things catch on.
    For example, why do we have to assume that because someone wants a “solar powered car” that the solar panels must necessarily provide 100% of the energy needed for daily driving?
    Does the fact that the Volt only uses batteries for 40 miles mean it is not worthy of consideration as an EV? Maybe to the purists, but for the average practical person, the Volt is an EV.
    According to my calculations, if we were to impregnate the bodywork of a typical EV with thin film solar panels and then park it out in the sunshine all day, then you’d be able to drive 10 km on that captured energy. For someone who commutes 20 km a day, that is significant. For someone who only uses the car every couple days, that may cover virtually all their needs. The rest of the energy comes from … plugging it in. But this 10 km a day would provide a lot of independence in the case of a monetary system collapse, gasoline rationing, or during prolonged blackouts when the economy falls apart. In this case, you wouldn’t be commuting to work anymore and that 10 km a day may be enough to get you by.
    As to the economics of this, of course it’s not terribly competitive with gasoline. But then again, neither were laptops competitive with big desktop systems when they first came out. If people didn’t believe in the idea, and if they didn’t throw some extra money in to support laptops initially, then they never would have developed. Now, an iPhone is more powerful than a big desktop system from 15 years ago.
    And with economies of scale, if auto manufacturers really started ramping up solar panel impregnated bodywork, then it might at some point become cost competitive.
    And the other thing wrt economics, what’s not often mentioned is the tremendous subsidies that fossil fuels get in relation to alternatives, so therefore any off the cuff comparison with say the price of gasoline today isn’t really valid. The whole $1.5 quadrillion derivatives market is a ponzi scheme designed to artificially suppress interest rates and keep the purchasing power of the dollar up (which is the same thing as suppressing prices, and the central driver to all other commodity prices is oil — therefore the purpose of the derivative scheme is to suppress the price of oil). Therefore, with an EROEI of 10, solar panels would fare much better in relation to oil if the market manipulation was removed.
    I am ordering a Leaf, hopefully get it in a few weeks. I am tired of getting sick from people sneezing on the bus, it is not worth it. I can’t wait.

  11. Theories are great. So here’s what actually looks like in reality. http://evalbum.com/1068. The solar power worked out at around one mile of driving per 2 hrs of sunlight. Only 3 real pluses. 1) lower depth of discharge on the batteries which translates into longer life expectancy. 2) If you run out of power, park it, and walk home. Come back the next day, no need for a tow. (yes I had to do this once). 3) Lots of attention, if your into that sort of thing.
    PS I retired the truck and sold the solar panels they are now here http://evalbum.com/3045.

  12. Thanks for the great post! I have a question about storage. With light rail vehicles I’ve seen several offer an ‘energy storage’ system like Bombardier’s MITRAC, which their brochure describes as a series of capacitors. Why capacitors and not batteries?

  13. Great article! I’m trying to turn some of the UCSB physics grads onto your blog because I think it’s only gotten better since you started, and crunching the numbers makes for a fun exercise.

    I was wondering about the potential use of electric cars for long distance travels. It seems to me that electric cars are great for commuters, but not too great if you want to go much further than that (conceivably one could want to drive from San Diego to LA and back, or about ~200 miles round trip in a day). The solar panel on your car seems like it might be a partial solution, but do you think there are others? I’ve heard ideas of battery stations that would swap batteries when they’re low or perhaps you could carry some additional juice if you were planning a longer trip. Perhaps the answer lies in quick charging batteries so that a small (~hour long) wait would be sufficient to make another leg of the trip. I just don’t see electric cars being anything but commuter cars (which, in fairness, is good enough for most daily activity). Or does the answer lie in gas-electric hybrids like the Volt (which I think may be the most viable solution, if only the electric could get you more than 40 miles!)?

  14. First off, the business as usual paradigm, that insists on something similar to a ICE powered passenger automobile is for all intents and purposes a very dead horse! It really is time to stop beating it… it ain’t getting up again!

    Having said that there are almost certainly going to be niche applications for solar powered EVs in our collective future.

    Personally I imagine vehicles such as solar powered electric assist velomobiles will become more ubiquitous. We don’t have a shortage of energy to power them. What we have at present are very unrealistic expectations. We’ve become a very spoiled bunch of crybabies.

    Now everyone, go to the blackboard, pick up your chalk, and write “paradigm change”, one hundred times. Or there will be no soup for any of you!


  15. Tom is overzealous, as usual, in his love for solar panels and other techno fantasies.

    The question that isn’t answered in this article, or any other discussion of so-called “renewables” for that matter, is this: If this thing [insert favorite techno-boondoggle here] is as cheap/easy/efficient/practical as you say, then why are we not doing it now?

    Solar PV is not a new technology, and electrical motors are most certainly not new. Why then do we spend billions each year on liquid fuels (no small part of which is sent to foreign suppliers)? Trillions more in blood and treasure on foreign wars to secure oil supplies? God only knows how much more in lost natural capital, environmental disasters (deepwater horizon, exxon valdez), climate change, and other planet killing “externalities”?

    Oh, that’s right, because batteries and PV panels don’t grow on trees. Meaning there is a very good reason those things are so expensive, and will not be getting any cheaper or more abundant as the fossil fueled industrial infrastructure they are completely dependent on for their very existence gets ever scarcer and dearer.

    What does grow on trees? Just an extremely elegant solar collector that is self-fabricated, self-repairing, uses only those materials that are available immediately on-site, and is one hundred percent recycled. A truly miraculous product derived from billions of years of trial and error.

    Humans and their cousins have thrived for millions of years on that system of solar energy and we will do so once again, but it probably will not be voluntary, and it most certainly will NOT involve driving cars.

  16. The golf cart idea is interesting. How heavy duty (speed wise) could you scale up a golf cart and keep it solar only?

    People have mentioned mass transit, which has two major problems:

    1) Availability: It leaves on a certain schedule, not when I want.
    2) Limit Access: Trains only go go specific areas, I need buses to get two and from the train.

    The first is still a major problem, but could a golf cart inspired minicar help fill the niche of the second? I drive this new minicar from my home to the train station, and another minicar to my place of work.

    You could even put security in the cars so that when they are plugged in, they make sure they are in their proper area, or they lock down (to try and curb theft).

    • Only mass transit as it exists today leaves on schedule. I can imagine – at least for larger cities – building a system where people use their computers or mobile phones to request a transport from A to B. The system would accumulate these requests and dispatch buses or smaller cars accordingly. I imagine that because there are some very frequented routes, there would be enough requests at any 10-15 minute period to make it sensible to send a bus to A and make it stop only at B and those places on transit that some other people requested.
      The system would be able to optimize the number of changes, size of the buses, etc. And if you think waiting 10 or 15 minutes would be too long, think about how much time you waste in traffic jams that would disappear if there was only centrally optimized trafic. And how much time you waste on finding a parking place.

      As to limited access, I think your solution is good – “the last mile” could be done by bicycles/solar golf carts/whatever.

  17. For transportation, I like the idea of combining solar with human power – i.e. cycling. I’m imagining occupants directly powering the wheels when the vehicle’s accelerating/cruising, and charging the battery when stationary/braking. It seems each occupant contributing 100W could reduce the required solar panel area by 0.3 m^2. Not fantastic, at least when the number of occupants is small. Still, it might reduce the required battery size, as more ‘instantaneous’ power would be available for acceleration. Having the ability to power the vehicle by cycling would reduce the risk of getting stranded on an overcast day.

    Presumably the math works out better as you improve the vehicle’s person per frontal area ratio, e.g. trains. If every occupant on the train were cycling, would a solar-powered train be feasible? I’d also like to see the math computing the steady state velocity a freight train covered in solar panels.

  18. A car that goes 10 miles per day (the golf cart example) makes no sense. A bicycle would be just fine.

    • Thing is, most people don’t want to exercise for their daily commute. It might do them a lot of good, but ultimately a golf cart esque cart would gain more public support than requiring people to ride a bike.

      Even if the bike was completely electric and required no human power, you can’t carry anything on it (even a briefcase would be an issue) and you would be exposed to the elements.

  19. Why is no one suggesting we go back to the horse and buggy? Makes sense to me. And it runs on fuel converted from the sun. Win all around.

    • At the turn of the last century, cities all over the world were being buried in horse manure, not to mention the logistics of managing this transport fleet. The problem was becoming so bad it was impacting further growth and provision of services. The fossil fueled transport revolution at the time was hailed as “miracle” cure for this escalating problem. And remember this was in a time of far smaller global population. If you think the current concerns about bio-fuels is intense now, imagine what it would be like with horses as the biofuel motors.

    • If you do the math and come to a conclusion that using biofuels to power cars is not feasible, the same applies to horses. I would think that cars are or could eventually be more efficient in converting energy into movement than horses, so arguments against powering all cars with biofuels would apply even more to horses.

  20. Hi Tom,

    I like this post as it addresses a common complaint of mine: people never ‘do the math’ with solar powered cars – rather they just assume the combination is obvious but we stupid engineers are incapable of ‘thinking outside the box’ to implement it.

    However, I think when you say “I want a solar powered car” you should also have addressed the psychology of wanting a solar powered car. People who want a solar powered car don’t necessarily want a car with solar panels on the roof, they just think they do. Actually they want a car that runs on solar electricity (probably they actually just want a car that is environmentally benign, but lets not go there!). Acheiving this by putting the solar panels (inherently low power density, moderate energy density) on the roof of the car (very high power density requirement, high energy density requirement) obviously makes no sense.

    More importantly, it was never a good idea in the first place! If you invest in solar panels, you should plan to have them in the sun as much as possible. Not in a garage, not in a parking building, not in a tunnel, not under a tree.

    So when we talk of solar powered cars we shouldn’t think of monolithic integrated solutions comprising solar panels on a car, but rather of system level solutions involving electric cars and grid energy with a high solar content. A solution which vehicles such as Nissan’s Leaf show are already a practical reality, complicated only by the continued high (but rapidly reducing) cost of solar PV energy and vehicle batteries.

    So – solar cars ARE a reality. We just need to look at the big picture.

  21. Reading your post again I realise this was your conclusion towards the end anyway – my bad. Only thing I take issue with is the suggestion that using grid tied PV to acheive the goal is less satisfying – if anything it’s more so because it makes you part of a community working to acheive a solution, rather than an island of sustainability among a sea of pollution.

    And in fact I don’t see any issue with outsourcing your clean energy – if you buy a leaf and tick the ‘renewables only’ box from your electricity provider then for my money you’re being as sustainable as someone with PV on their roof. Still issues of whether this would work if everyone did it, but that’s true with either solution.

  22. So, at the end solar cars are just a story of science fiction. Electric cars are still to far of being affordable for the American regular family, solar cars are still a dream from far far away.

    • A car like the Volt, even at a sticker price of $45k, is not unaffordable once you “do the math”.

      Sticker Price: $45.5k
      Tax Credit: $7.5k
      Gasoline saved over 5 years (for me): $11k
      Equivalent to a gas car costing: $27k

      The average sale price of a new car in the US is around $29k. So, the Volt is CHEAPER than the average car over 5 years. Keep it longer, or assume gas prices go up, and it gets even better.

      • Maintenance is also a significant area of cost savings. Even hybrid owners already find that brake pad maintenance largely disappears do to the regen braking taking much of the load, though its rarely mentioned, I expect, due to out-of-sight, out-of-mind. Fleet EV managers seem particularly aware of the potential maintenance savings:

        “On an equivalent 100 mile-per-day diesel vehicle, we spend roughly $900 per year in preventive maintenance – oil changes, filter changes, anti-freeze adds, and eventually transmission oil changes. With the electric vehicles, we take that down to $250 per year.

        The electric trucks are only equipped with four grease fittings and no engine or transmission oil. The truck must still be taken to look at brake lines and other wear components that may be cracked. Overall, there is virtually nothing that goes wrong with these things.” – Staples vehicle fleet manager


    • Uhhh… I’m not sure whether you’re trolling or not, but for now will give you the benefit of the doubt.

      It’s true that they’re not yet the highly affordable reality. But that doesn’t mean they’re anything like as far away as science fiction would suggest.

      Electricity from solar PV is now being forecast as cost competitive with coal in good locations within the next decade. Wind already is, or close enough.

      Batteries are improving so rapidly I think that the general understanding of where they’re at is way off the mark. Batteries that last thousands of cycles (read, hundreds of thousands of kilometers/miles) and can feasibly be integrated into vehicles at <$250/kWh are already beyond lab stage and moving into early production.

      There's still debate on PV, I agree. But for the bulk of (at least urban) mobility I'd say it's not a question of if electric cars take off but rather just whether it's 5 years from now or 10. My money is on 5. Or less.

      The next question I'm really interested in the answer to is just how much better we can do than the cars of today by lighter weight, active crash avoidance, mobility on demand, and smaller vehicles for most trips. Cities of the future could be really beautiful places to live if we don't totally f*ck the climate and social structure in the meantime. Next year will bring the Twizy. I like in hope GM's EN-V will get out of the prototype shop eventually. These are game changing concepts.

      • You bring to mind something pretty much on-topic that perhaps somebody here might have some ideas about.

        I’ve got a PV array on my roof with a good amount of excess capacity. I’d like to take advantage of that capacity with a car that has the following minimum specs:

        * Seats at least two with room for overnight luggage. The classic (aircooled) VW Bug would be luxurious; the Honda Insight would be just fine.
        * Capable of cruising at 65 MPH with a maximum speed of at least 75. Less than that and you’ll get run over on the freeways here, and a car that can’t go on the freeways is useless — that’s what the bicycle is for.
        * Minimum real-world range, fully loaded in the Arizona summer with the A/C on should be not less than 100 miles. That is, I should be able to drive to my friend’s home 40 miles away on the other side of the Valley and not have to even begin to worry about whether or not I’ll be able to get back to my own home without plugging in. And do I have enough charge to stop for groceries on the way back?
        * Price should be well under $20,000 — after all, any petrol-based car with those specs would be, and the “early adopter” premium is only worth so much. I’m not going to pay luxury car prices for a glorified golf cart. Used is okay, though.

        I don’t think that’s too much to ask for, but I’ve not seen a hint of anything like it. Any suggestions?

        I’m sure there will be something like that in the next few, or at least several, years…but I’d rather not have to wait quite so long….



        • While there’s nothing you can buy today that exactly fits the bill, the Leaf is actually very close to what you’re asking for here. I’d say that as long as you’re prepared to pay the $20k (rather than a lot less) you’ll be able to get what you describe in 2-3 years max.

          In the meantime, just sell your solar PV energy back to the grid. Sure, it means your PV panels aren’t displacing oil… but they’re displacing coal instead which is about as bad.

          On the range issue, as fast chargers become more widely available this will really stop being a concern for commuter driving (with the Leaf you can already get a 20mile top-up in about 6 mins using existing commercially available systems). Long trips are still a problem – With a top end Tesla (especially with battery swap models) less so… but taking care of commuter driving first will be a huge step forward.

        • One extra point on price – have you driven or ridden in a Leaf? I tried one out on the road back in March – it’s a REALLY nice car in terms of interior, build quality, comfort, and handling. I’d compare it to a top end (non-turbo) Subaru Impreza, not a golf cart, and not anything like some damn Chevy Aveo… 🙂

  23. “First, let’s examine the requirements. For “acceptable” travel at freeway speeds (30 m/s, or 67 m.p.h.)”

    I read this statement and the first thing that comes to my mind is BAU. Less than 20 years ago and after the shortages of the 70’s the speed limit was 55mph. If this country went back to an enforced 55mph it could save a million barrels of oil a day and electic vehicles would have a chance to complete with gas & diesel.

    55 is the first step to really addressing the worlds BAU energy problems. Until then it’s all really just a bad joke.

  24. A solar powered car makes much less sense than an electric car powered by a solar array on the home. I have net metering, so if I make more than I need I run my meter backwards. Since April my house has powered itself AND my transportation to work.

    • I made the same argument above, and on review I felt that was actually Tom’s point towards the end of his initial post anyway.

      However, there is one critical difference – Solar panels on the car being use for instantaneous propolsion requirements would avoid the battery problem and also increase source->wheels efficiency. So it would be nice if you could make it work.

      But in all realistic usage cases it doesn’t work, so I’m perfectly happy to accept grid solar and electric cars as an alternative.

  25. Here is a video about my homebuilt electric car and my solar powered roof:


Comments are closed.