Energy Collective blog power policy climate - the conversation happens here

Tuesday, December 13, 2005

Who Needs Gasoline? - The Incredible Potential of Plug-ins and EVs

How would you like to power the entire U.S. fleet of cars and light trucks using only half the oil we currently use? What would you say if I told you we could even do it without using a single drop of petroleum?!

Today we're going to talk about how we could do just that, right now, using available technology: electric vehicles and plug-in hybrids. It all comes down to how efficienctly we want to utilize our petroleum (and indeed energy in general) for transportation...

The Current Picture - An Inefficient Waste

Currently, our transport needs are met by petroleum refined into gasoline or diesel and then burnt in internal combustion engines. This process is not very efficient overall. Here's how it breaks down, from the barrels of oil to the energy at the wheels of our cars:

  • Engine combustion efficiency: ~35% (gasoline engines are about 20-30% efficient and diesels are ~40% efficient so we'll take 35% as a combined average to represent the two)

  • Engine mechanical efficiency: ~70% (efficiency of tranmitting mechanical energy of pistons to the driveshaft and transmission)

  • Transmission efficiency: ~70% (efficiency of transmitting mechanical energy of driveshaft and transmission to the wheels

  • .35*.7*.7 = Total barrel (BBL) to wheel efficiency = .1715 = 17.15%

    [Note: This doesnt take into account the efficiency of refining the crude oil into gasoline or diesel (which I don't know, does anybody have any idea?) which would drive this efficiency down even farther.]

    Here's how this breaks down in terms of oil consumption: according to the TDB (p. 1-18) we currently use about 13.2 million barrels of oil per day (million bbl/d) for transportation. 56.6% of this or 7.4712 million bbl/d go towards running the U.S. fleet of cars and light trucks (TDB p. 2-1). At 5.8 million BTU energy content per bbl of oil, that's 43,332,960 million BTU per day of energy used by our cars and light trucks. At 17.15% overall efficiency, only 7,431,602 million BTU of that energy makes it to the vehicles wheels.

    So, is there a more efficient way to get the energy from the barrel of oil to the wheels of our cars? Fortunately, yes there is.

    Electric Vehicles - Who Needs Oil?

    What would it look like if we used all that petroleum to generate electricity to power an electric vehicle fleet? Something like this:

  • Generation of electricity (burn petroleum in a combined-cycle plant): ~50%

  • Transmission of electricity: ~90%

  • Battery efficiency (lithium-ion): ~90%

  • Electric Motor Efficiency: ~90% (notice how much more efficient an electric motor is than an internal combustion engine!)

  • Transmission Efficiency: ~90% (an electric transmission is also more efficient than one for an ICE)

  • .5*.9*.9*.9*.9 = BBl to wheels efficiency = .3281 = 32.81%

    That's an improvement of 191% or almost double the efficiency of an ICE fuel pathway.
    Right there, just by burning our petroleum in efficient combined cycle plants, rather than inefficient gasoline and diesel internal combustion engines, to create electricity and run an electric drive-train vehicle, we could cut our oil consumption for cars and light trucks almost in half. We would only need 52.28% of the oil we currently use, for a savings of 3.565 million barrels of oil per day!

    [Note: this example doesn't take into account the efficiency of refining the crude oil into something suitable to burn in a combined cycle plant but I didn't take into account the refining of gasoline for ICEs either so this should be a wash. More importantly, this doesn't take into account the gains of regenerative braking since I'm not sure how much they would be. I would imagine that they are not insignificant however and if anyone has a good estimate for me, I'll plug that in and see how that changes things]

    OK, but I promised you we wouldn't need a drop of oil and so far I've only cut our oil consumption in half. We've still got a long way to go right? Wrong. The beauty of using electricity as your primary fuel is that it is not dependent on petroleum to generate. Unlike gasoline, electricity can be made from a myriad of sources, some renewable and others not, all of them producable from within the United States (unlike oil which requires us to import almost 2/3rd of our oil needs and is never renewable). That means we could potentially eliminate our dependence on oil for our car and light vehicle fleet completely if we were to transition to an all-electric fleet.

    But that would require a whole ton of additional capacity wouldnt it? We were going to use that 3.9058 million bbl/d of oil to generate 11,326,936 million BTU in our combined cycle plants or the equivalent of 3,319,925 MWh per day of electricity (1 million BTU = .2931 MW). To meet the needs of our EV fleet would thus require 138,330 MW of added capacity (1 MW = MWh/d / 24 hours). Put another way, that would require 138 new power plants (average size 1000 MW) to be built. According to the EIA, total U.S. generating capacity in 2004 was about 1,000,000 MW or 1,000 GW. Powering our EV fleet would thus require us to add another 13.83% to our generation capacity with the addition of over 138.3 GW additional of generating capacity. Or would it...

    The Zero-Oil Solution - Using What We've Already Got

    Peak demand during the day is generally about twice that of nighttime demand. That means half of our current generating capacity is already sitting idle all night. If we were to charge our EVs overnight, we could utilize this idle off-peak generating capacity instead of building new power plants. Let's see where this would get us:

  • Total U.S. generating capacity in 2004: ~1,000,000 MW

  • Percentage of generating capacity idle at night (let's say 10:00 pm - 6:00 am): ~50%

  • Total U.S. generating capacity idle at night: 500,000 MW

  • Total energy output of idle capacity each night (500,000 MW * 8 hours): 4,000,000 MWh

  • Total energy output of idle capacity each night: 13,647,219 million BTU

  • Electrical needs of EV fleet: 11,326,936 million BTU

  • That means we wouldn't need to construct any additional capacity and we could power the entirety of our car and light truck transport fleet from existing idle off-peak generation infrastructure, essentially eliminating the use of oil to power our light vehicle fleet. [Some of that existing generating capacity we would be using at night would be powered by petroleum still... so I guess I lied, we would use a bit of oil but we wouldn't have to, these plants could easily be replaced by greener sources like wind and solar, the possibilities are wide-open now that we are using electricity...]

    There we have it, we can power our entire light vehicle fleet with electricity from off-peak power plants, eliminating the need for 7.471 million barrels of oil per day and without constructing a single new power plant. If that isnt impressive, you're not paying attention.

    The Added Freedom of Plug-ins - a Suitable Compromise

    OK, so about now is where all those folks who complain about the limited range of electric vehicles start chiming in. What do we do when we want to drive to Grandma's house for Thanksgiving or take the kids on that roadtrip to Disneyland or head down to Mardi Gras for Spring Break? We don't want to have to keep two cars in the garage, one for daily trips and another for those longer road trips where an EV's limited range wouldn't cut it?

    Well, these are all valid complaints. Although it seems like many people could afford to have two cars (many people already do), most would be reluctant to do so, especially those used to the freedom of a gas-guzzler - you can pull in to a gas station anywhere and in five minutes be back on the road and free to drive another 300 miles.

    Well here's a compromise: plug-in hybrids. That is, a hybrid electric vehicle that can be plug-in to charge and can run in all electric mode for 20-40 miles (like an EV) but can also use gas or diesel on extended trips (like an traditional ICE).

    This means there are essentially two fuel pathways for this vehicle. The first is the electricity pathway that would be the same as for an EV. The second would use a gas or diesel generator onboard the car to burn fuel and generate electricity to run the electric drive-train when the batteries are dead [this is called a series-parallel hybrid system and is found in the Mitsubishi Concept-CT MIEV for example].

    If we assume an all-electric mode with a 40 mile range, we could probably safely say that at least 2/3rds of our trips could run in all-electric mode. According to the Bureau of Transporation Statistics, the average person in the U.S. drives an average of 40 miles per day so most daily trips could be taken using only the electric charge. As we discovered above, this power could easily be provided by existing idle off-peak generation capacity if the plug-in vehicle was charged overnight. That means that just like the EV discussed above, 2/3rds of the driving done in a plug-in could be done without any oil!

    So, what about the remaining 1/3rd of the time when the plug-in runs on its gas/diesel generator? Here's what the efficiency of that fuel path looks like:

  • Engine efficiency (diesel gen-set): ~40%

  • Generator efficiency: ~80%

  • Electric motor efficiency: ~90%

  • Electric transmission efficiency: ~90%

  • .4*.8*.9*.9 = BBL to wheels efficiency = .2592 = 25.92%

    [Again this doesnt take into account refinery losses or gains from regenerative braking]

    Notice this is more efficient than the ICE fuel pathway, which is why we use the diesel gen-set to generate electricity and run the electric drivetrain rather than using it to run an ICE engine and drive-train (the latter is what they do in all hybrids currently on the market - Toyota's 'Synergy drive' for example features a split drivetrain that uses both an electric motor and an ICE engine).

    What does this mean in terms of oil consumption? Well, we would need to get only 1/3rd of the at-wheels energy from the diesel gen-set or 2,526,745 million BTU. With 25.92% overall efficiency that would mean we would need 9,748,244 million BTU of petroleum (i.e. diesel) fuel or the equivalent of only 1.680 million barrels per day. Thus we would need only 22.5% of the oil we currently consume for our light vehicle fleet for a net savings of 5.79 million barrels of oil per day. This would eliminate 28.38% of total U.S. oil consumption!

    So there's the compromise: we don't entirely eliminate our need for oil to run our transport fleet, but we do cut it down to less than one quarter of current consumption and eliminate over 28% of our total oil consumption at the same time. Doing so gives us the freedom and range of a traditional ICE engine while using less than a quarter of the oil.


    Clearly using petroleum in internal combustion engines is not the most efficient way to run our light vehicle fleet. Both electric vehicles and plug-in hybrids offer significantly better options. They also allow us to use electricity, giving us the freedom to power of fleet with a variety of energy sources, not just petroleum. Additionally, the United States currently has the existing generation capacity sitting idle every night to power our entire transport fleet if they were EVs or to fully charge the batteries on all of our plug-in hybrids.

    Here's the summary:

  • Traditional Internal Combustion Engine: BBL to wheels efficiency - 17.15%; total oil consumption - 7.47 million bbl/d - % below baseline - 0%

  • EV with electricity from petroleum-fired combined cycle plants: BBL to wheels efficiency - 32.81%; total oil consumption - 3.9 million bbl/d - % below baseline - 52.28%

  • EV with power from existing idle off-peak generating capacity: BBL to wheels efficiency - n/a; total oil consumption - 0 bbl/d (excepting petroleum in existing generation mix); % below baseline - 100%

  • Plug-in with batteries charged from existing idle off-peak generating capacity and extra power from on-board diesel gen-set: BBL to wheels efficiency - batteries: n/a gen-set: 25.92%; total oil consumption - 1.68 bbl/d (excepting petroleum in existing generation mix); % below baseline - 77.5%


    Jesse Jenkins said...

    Actually, Frosty, using hydrogen is less efficient, quite a bit less efficient in fact. Here's how the efficiency breaks down for hydrogen (we start with electricity from the idle off-peak generating capacity of our existing plants):

    -Electrolysis of water to produce hydrogen: 70% efficiency
    -Liquidification, transport and storage of hydrogen: 80% (this is a guess but I can't imagine it would be better than 80% as liquidification is energy intensive)
    -PEM fuel cell to create electricity: 40% efficiency
    -Electric motor: 90% efficiency
    -Electric transmission: 90% efficiency

    .7*.8*.4*.9*.9 = electric generation to wheels efficiency = .18144 = 18.144%

    Now, obviously that is really low overall efficiency (if you add in the power generation step -lets say ~40% - then we end up with a total efficiency of only 7.26%).

    The simple fact is, using electricity to create hydrogen to then create electricity again is always going to be (significantly) less efficient than simply using the electricity in the first place as we do with an EV or PHEV.

    Here's what the efficiency of the EV (or the battery pathway of the PHEV) looks like in comparison (again we start with electricity from idle off-peak generating capacity):

    -Transmission of electricity: 90% efficiency
    -Battery: 90% efficiency
    -Electric motor: 90% efficiency
    -Electric transmission: 90% efficiency

    .9*.9*.9*.9 = electricity to wheels efficiency = .6561 = 65.61%

    So no, hydrogen would not be more efficient. In fact, an EV powered by idle off-peak generation capacity would be over 3.6 times more efficient than using that off-peak capacity to generate hydrogen for hydrogen fuel cell vehicles.

    (This is why I didn't use hydrogen in my examples)

    Jesse Jenkins said...

    I've been informed that perhaps my numbers for the ICE engine and diesel gen-set for the PHEV might be a bit high. I may adjust these latter and see what that does to the picture - it will make the EV even better compared to the ICE and PHEV and the PHEV and ICE will likely stay the same relative to each other.

    Also, while I've indicated that significant generating capacity lies idle at off-peak times, I must note that that does not necessarily mean there is fuel available to fire those plants at night.
    However, I bring up idle off-peak capacity in order to illustrate a point: i.e. it would not take significant infrastructure additions to generate the electricity to power an all EV (or PHEV) light vehicle fleet.

    In reality, some new capacity would likely have to be built as existing natural gas plants in particular would probably have a hard time finding fuel to run at such a high rate 24 hours a day (many of them are 'peaker plants' and are only designed to run a small portion of the time, i.e. when prices and demand are high). Also, we wouldn't necessarily want to power our EV fleet on our coal plants, which would contribute to significant air pollution and greenhouse gas emissions.

    However, as I said in the post, using electricity to power our fleet gives us the freedom to turn to a number of generation sources. We could construct new wind farms to meet a portion of the extra needed capacity, for example, and if all else fails, we can always burn some of the oil we are no longer using in combined-cycle plants. This couldn't be worse than what we already do - in fact it would be much better as we would need to burn much less oil, combined-cycle plants are much more efficient than internal combustion engines and a centralized location would make emissions controls significantly easier.

    Heiko said...

    Here's the estimate you requested.

    Anonymous said...


    Where is the Professor Fate cheer?

    Where is the Frank Axiom?

    Where is the plug for that PHEV site, eh?

    Jesse Jenkins said...

    Thanks to whoever the anonymous coward who refered me to the After Gutenburg blog. There's a lot of discussion of plug-ins over there. The post the comment refers to points out several useful things:

    1) The new Hybrid Consortium for Plug-in development has a website now. Check it out for more on plug-ins.

    2)The technology for plug-ins is not some far-fetched technology, nor is it even still a few decades off (like hydrogen fuel cells or perhaps electric vehicles with the range and flexibility demanded by US consumers). It features hybrid technology already in mass deployment (think Prius) with the addition of a larger battery and a charge regulator. CalCars has converted a number of Prius's to plug-ins already and it appears that Daimler-Chrysler is also using its new 2006 Dodge Sprinter as a platform for developing what would be the first commercially available plug-in hybrid. The Sprinter plug-in diesel-electric hybrids are currently undergoing commercial trials in the U.S.

    Anonymous said...

    Great post Jesse, keep 'em coming.

    The Boiled Frog said...

    I was wondering where you got the value of only 10% transmission losses of electricity. I will do my best to find the information, but I believe the Rocky Mountain Institute (RMI) estimated transmission losses as high as 75%. So for every one watt lightbulb lit at home, four watts of power are being generated. That's why one of the most important thing consumers can to do to decrease our use of energy is conservation at home. Use 25% in your equations (the worst case scenario) for transmission, and the results are quite different. Like I said, I don't have a link at this time, but I will do my very best to get this to you ASAP. I don't want to be too negative for my first comment on yor site, I think this one of the best sites in the blogosphere, keep up the excellent work!

    Roger, Gone Green said...

    For most uses an all electric works fine for me; no need to be able to drive 3000 miles non-stop (other than for five minute refuels). I drove an early Saturn electric for a couple of weeks and it was AMAZING.

    In any case, to take it one further, there is a local company that will build a "solar carport", a charging station at your home for your car that runs off solar.

    I was all set to buy an Electric RAV4 when the Bushies sued to kill California's higher fleet emissions standards and toyota "recalled" them . . .

    Sinus said...

    75% electric transmission losses?

    That's ridiciluos. Transmission eff ranges between 80%-95%

    Lower (or higher) is considered an exception.

    Jesse Jenkins said...

    Sinus, that's what I've read as well, which is why I used 90% as my figure for my calculations.

    Frog, I've got to assume you misread something there because 75% transmission losses is rediculously high. Maybe 75% efficiency but that would still be quite a bit below normal.

    Unless you've got the source of that figure, I'm gonna stick with my figure of 90% as a reasonable estimate for transmission losses.

    Jesse Jenkins said...

    The battery infranstructure is an issue but one that should be able to be resolved. I will certainly look into this more as I continue my thesis-related research. However Vincent, it most certainly won't be lead-acid batteries in these plug-in hybrids or EVs (too poor energy and power densities) but most probably lithium-ion batteries and ultra/supercapacitors.

    Jesse Jenkins said...

    I haven't seen the e-Traction site before. Thanks for the tip, Vincent.

    Anonymous said...

    Just came across your pre-xmas discussion

    US Energy transmission/distribution losses 7.2%

    Anonymous said...

    Wikipedia has several references analyzing energy losses under their Losses section on Electric power transmission.

    Anonymous said...

    The problem is that oil (and gasoline) allow storage in cars of vast amounts of energy at very low cost. The current state of battery technology doesn't even come close, and it may never. This may be more of a marketing challenge than a technical problem, but you cannot expect to replace the gasoline infrastructure with the electric grid. Where are you going to store the energy at believable costs?

    Jesse Jenkins said...

    Tom, the battery technology isn't really as bad as folks often think it is. Rapid progress is being made on lithium-ion batteries for vehicle applications and the energy storage capacity is starting to get high enough to provide reasonable ranges for EVs. (Check out The Energy Blog's energy storage and battery archives and you'll see what I mean)

    Additionally, it makes the most sense to start with plug-in hybrids first which need smaller batteries with less storage capacity than full EVs. Remember, you only need to get a range of 20-50 miles in electric mode - the rest comes from gas still. Even that though gives a tremendous savings in energy use as the post indicates. And as demand for plug-ins fuels further development of advanced battery technology, batteries suitable for full EV applications would likely be developed before too long.

    "Where are you going to store the energy at believable costs?," you ask: in the batteries onboard the vehicles, of course. As you seem to be hinting at in your comment, the energy grid does not currently have much storage ability. Adding millions of batteries in plug-ins and EVs can be what provides that missing storage ability for the grid. Charge overnight when loads are low and run the car during the day. Helps level demand.

    If smart grid technologies are implemented (i.e. the grid plus networking/communications technology), you could also plug your car in during the day to act as smart storage capacity for load leveling the grid and even earn some money or free charge off of it.

    Finally, for stationary energy storage coupled to the grid, flow batteries (see here and here) would make a lot of sense. 'Gas' stations could become electric fueling stations for quick charges to top off the battery during the day (remember this is just to supplement overnight charges) by installing flow batteries. They too could be charged overnight to be ready to top off the charge on electric vehicles during the day.

    Anonymous said...

    I think plug-in hybrids are a really great idea, but what I was trying to point out is that there are a few technical problems in the way. Replacement batteries for a Prius cost about $3,000 today. This is a metal-hydride battery. You can do better in the open market, paying about $300/KWh. A lithium-ion one costs about twice that. And this battery contains about 1KWh, as opposed to the gas tank, which contains 436 KWh, and can be refilled at any gas station on the road, within minutes. Battery technology is going to improve, but it has a long way to go.

    This constraint is going to be there as long as buyers want to be able to jump in the car and drive to Timbuktu. How often do people really need the capability? Not very often. In reality, a modest increase in battery capacity is capable of capturing most driving, but this is mostly a marketing problem, and the challenge is to alter human perceptions. I'd rather have a technical problem

    Charging that extra capacity at night makes a world of sense, too, but even though most of the system is idle, fuel is going to be consumed, and a lot less efficiently than when the plant is running at capacity. The price may be dominated by capital costs, and so if anything, electricity in the dark of the night is going to be cheaper, at least until a lot of people start charging at night. The car needs to be smart enough not to charge an extended battery with expensive electricity, and I suspect this is going to make regeneration impossible.

    Jesse Jenkins said...

    "The car needs to be smart enough not to charge an extended battery with expensive electricity, and I suspect this is going to make regeneration impossible."

    Care to elaborate on that last part? I'm not sure what you mean but I don't see any reason why making the batteries 'smart' enough to only charge when prices are cheap/demand is low would prevent regenerative braking...

    Anonymous said...

    "The car needs to be smart enough not to charge an extended battery with expensive electricity, and I suspect this is going to make regeneration impossible."

    What I meant by that was that although one might fill one's battery with cheap power when charging at home, regenerative braking is only going to recoup power that would otherwise be wasted. You're going to need that power to accelerate. Inefficiencies in the system and aerodynamic drag are going to, slowly but surely, use up one's stash of cheap power. Moreover, there is not only a trip to work to consider, but the trip home. And, when the cheap power is gone, one must use the output of the ICE, which is much more expensive than the power you got from the power net. I suppose that it is possible that a car will arrive at work, bearing a load of cheap power that the driver can then sell, but given the projected state of the art in batteries, that seems unlikely. It also seems unlikely from the point of view of human nature that someone would sell a resource that they might need unless they are very sure that they have a surplus in hand. Actually controlling a battery so that it won't recharge it's losses using the ICE seems like a pretty easy computer stunt - I would be surprised if that was a problem. It's the other parts that seem problematic.

    Anonymous said...

    I don't understand this part:

    # Engine efficiency (diesel gen-set): ~40%

    # Generator efficiency: ~80%

    Isn't the generator part of the genset? Is the overall generation efficiency (diesel-in to electricity-out) only 32%?

    If first saw a serial hybrid on the cover of popular mechanics forty years ago. (Crazy guy had put an electric motor and a 20K-constant-RPM turbine generator in an Impala.) Seemed wildly sensible then, still does (well, maybe with a [turbo]diesel generator instead).

    Do the economics not work? Why don't they produce these?


    WattHead said...

    Yeah, you got that right. The total efficiency of the diesel engine and generator (i.e. diesel energy in, electricity out) would be 40%*80% = 32%.

    Yes, it's a perfectly sensible idea. Chevy's Volt concept uses it and Ford has a similar concept platform. See this post.

    The hold up seems to be commercialization and mass production of the batteries. GM says the Volt can roll out in 2010. Other companies - including a Chinese company, and several outside of Detroit start-up automakers - seem to think they can beat GM to the punch. It's only a matter of time now...

    FR said...

    There's at least one problem right off the bat. Plug-in hybrids are always going to be much more expensive than the cars we're used to, because they have to have both a very expensive battery *and* the gasoline combustion engine.

    Also, we'd need lithium-ion batteries for these; no other metal even compares to the performance of these. Is there enough lithium in the world to make many of these cars? I'm asking; I don't know.

    If those concerns are addressed, how long would it take to scale up the production of these?