The Mirai’s MPGe at 50 mph

The US Environmental Protection Agency (EPA) has estimated the combined city/highway miles-per-gallon-equivalent (MPGe) of the Toyota Mirai at 67 MPGe, and the range of its 5 kg tank as 312 miles.  Since neither of these numbers are terribly useful without knowing the speed and traffic conditions, I decided yesterday to get a better idea of how the MPGe depended on speed.

I’d already observed that the EPA’s 67 MPGe number was almost exactly reached on a trip on June 25 from the Saratoga station down US-101 to the Santa Barbara station, and that on average over the 1700 miles I’d driven the car so far I was getting 57 MPGe, due mainly to highway driving at the prevailing traffic speed of 70-80 mph on US-101 between Palo Alto and Pacific Grove.  What I didn’t know however was whether a higher MPGe was possible by driving at a steady 50 mph on country-road conditions unimpeded by stalled traffic and traffic lights.

So on the afternoon of Yellow Pig Day 2016 (July 17) I spent a couple of hours on various roads within 30 miles of Palo Alto with the cruise control set at 50 mph.   Being a hot day (85 F) I set the A/C to ECO and the climate control to 77 F, which was much more comfortable than with the A/C off.

The Mirai provides a number of gauges for assessing fuel economy, selectable using some of the buttons on the right of the steering wheel.   The one I found most useful (I’ll call it the TSF gauge) shows the Elapsed Time (HH:MM), Average Speed (MPH), and Average Fuel Economy (MPGe) for the selected trip meter.  There are four trip meters, A, B, P (my name for it), and ODO.  These can be selected by pressing the TRIP button on the steering wheel, and the corresponding miles-traveled is shown underneath the speedometer, independently of whether the TSF gauge is selected.  A and B can each be reset to zero by holding down the TRIP button on the steering wheel for two seconds.   P is reset to zero at Power-on, and ODO is the odometer.  For any of the four trip meters the product of miles-traveled and Average Fuel Economy on the TSF gauge gives the kilograms of hydrogen consumed since that meter was reset.

(It would have been more useful to have P reset automatically at fill-up than at Power-on, though resetting A manually at fill-up is almost as good provided you remember each time and don’t mind using up one of the two trip meters in this way.  With that trip meter, subtracting kilograms-consumed (as calculated above) from the tank’s capacity of 5 kg gives the remaining hydrogen.  I find remaining-hydrogen far more useful than the Mirai’s estimate of remaining range because the latter is based on the last eight fill-ups and is therefore not very useful when you can predict better than the car can what your likely driving pattern will be for the remaining tank contents.

After experimenting inconclusively on I-280 (too hilly to be informative), I settled on Foothill Expressway for my first experiment.  I started from Page Mill Expressway and ended a mile or so after passing under CA-85 where the 45 mph speed limit dropped to 40 mph.   After the first couple of miles the TSF gauge showed around 100 MPGe but then I started encountering red lights which soon knocked it down to the 70s.  Since the goal was to measure economy at a steady 50 mph I discarded that and decided that 50 mph on CA-85 would be easier to maintain.  So I went back to 85, and since the Saratoga hydrogen station was only six miles away I filled up there so as to start the measurement with a full tank, with the plan being to refill afterwards and compare the pump’s and car’s estimates of how much hydrogen had been used.

As usual I reset Trip Meter A at the station.  However to get a better idea of the MPGe at a sustained 50 mph I waited until I was on 85 and going at 50 mph before resetting Trip Meter B.

My plan was to drive to Gilroy and back, a round trip of 70 miles.  However at Cochrane Road, 21.8 miles, I could see stalled traffic ahead so I cut the trip short, took the Cochrane Road exit, and returned to the Saratoga station.

At Cochrane Road the MPGe was 99.6, which blew me away.  I figured the trip must have been downhill some of the way for the MPGe to be so high, and therefore expected the return to be lower.   So I was enormously surprised to see it actually increase on the way back, reaching 100.9 MPGe as I pulled into the Saratoga station 0.7 miles from the exit off CA-85.

My Mirai’s odometer is consistently showing 3% less distance traveled than what Microsoft Streets and Trips (one of the better navigators for PCs) claims. So going by S&T my real MPGe should be 104 MPGe. Not that there’s a big difference there when compared with the EPA’s city/highway figure of 67 MPGe, nor with my own result of 56.6 MPGe averaged over 1800 miles of lead-footed keeping-up-with-traffic driving since acquisition.

When I filled up on the return to Saratoga the pump showed 0.446 kg. The car judged the round trip at 46.0 miles (the 45.0 miles on the gauge plus 1.0 miles from the station to where I reset Trip Meter B on reaching 50 mph on CA-85), so the 100.9 MPGe figure means that the car thought it had consumed 46.0*100.9 = 0.456 kg, a difference of 0.01 kg or 1/3 oz (and .01/.446 = 2.2%) more than the pump’s reading. This difference is well within the pump’s claimed accuracy of 5%, but it’s only 0.2% of the tank’s capacity and therefore insignificant. I’ll start keeping an eye on the car’s MPGe number in future to get a better idea of car-pump agreement on longer trips.

On the social-interaction aspect of driving at 50 mph in the rightmost lane of a 65 mph freeway, I have only two observations. In the three hours I spent on these experiments, only one motorist expressed any arguably overt hostility and that was a taxi driver who briefly honked at me as he passed. (And for all I know he was just appreciating my PROTON POWERED license plate. Honking doesn’t offer quite the same explanatory bandwidth as Tolstoy’s War and Peace.)

What I did find however was a number of leeches that attached themselves almost to my rear bumper, each for several miles. My first thought was that this was how they complained, and my blood pressure went up a bit. But then it occurred to me that maybe some motorists are happy not only to drive slowly for better mileage but in the draft of another slow driver for even better mileage. This thought was reinforced by the observation that sometimes I had a caravan of several cars tightly anchored to me, none seizing any of the frequent opportunities to merge left into a faster lane.

So what does this mean for the Mirai’s so-called range? Well, if you believe Streets and Trips over the Mirai’s odometer, and you can avoid traffic jams and traffic lights, and that all 5 kg of the Mirai’s two tanks is available, and you aren’t experiencing major altitude changes, then by setting your cruise control to 50 mph and setting your A/C to a tolerable level and driving off into the sunset you should be able to cruise into the next hydrogen station after driving 520 miles. Or if you drive out and back from say Truckee or San Juan Capistrano with your cruise control at 50 mph you should be able to go as far as 260 miles (252 by the Mirai’s odometer until further confirmation) before turning around.

Suggestion to Toyota: a trip meter that resets at each fill-up. This could replace the one that resets each time you start the engine, which doesn’t seem terribly useful. Currently I use Trip Meter A for that but this entails resetting it at the station.

Posted in Uncategorized | Leave a comment

Carbon footprints

California awards Zero Emission Vehicles or ZEVs its coveted white sticker, entitling its solo driver to drive in the high-occupancy-vehicle (HOV) or carpool lane.  The state’s rationale is that CO2-emitting vehicles are contributing to a gradual rise in temperature over the past century or so that is expected by 2100 to have raised the planet’s temperature 3 degrees C above its relatively steady level prior to 1750.  The white sticker is one of the state’s incentives for car-owning individuals to reduce their so-called carbon footprint, the rate at which they are contributing to the planet’s rising CO2.

The state considers a ZEV to be a vehicle that does not emit CO2.  Battery electric vehicles or BEVs fit that description.  The first production BEV was built in 1884 by Thomas Parker of London.  They were popular early in the 20th century, albeit not for emission reasons, but soon lost out to the more economical, powerful, lighter, and longer-range internal combustion engine vehicles (ICEVs) based on fossil fuels such as gasoline and diesel,

The popularity of BEVs was revived a century later when their ZEV status was recognized as contributing less to global warming than ICEVs.  Over the past fifteen years BEVs have gradually gained an enthusiastic following, with price highly correlated with range: each additional $1000 in price gains about 3 more miles of range.  To within an accuracy of 10%, a $33,000 Nissan Leaf has a range of a hundred miles, the $67,000 low end Tesla Model S extends this to two hundred miles, and the $100,000 high end Tesla further extends it to three hundred miles.  Whereas luxury brands such as BMW and Mercedes are priced to offer only 2.5 miles per $1000, the price of the high end Tesla is primarily to pay for the larger battery needed to maintain the range-to-price ratio.

Deep cycling of a lithium-ion battery shortens its life.  To maximize battery life all BEV vendors therefore recommend keeping the state of charge in a range something like 30% to 80%.   This effectively halves the usable range between recharges.  For long trips one can ignore this recommendation but only at the cost of reduced battery life.  For commuters with a daily commute less than half the range and with little time off for long vacation trips this is not a concern.  However for fleet owners, taxi drivers, traveling salesmen, etc. limited range and especially the longer refueling time are serious restrictions.

A far more recent ZEV entry is the fuel cell vehicle or FCV.  Unlike BEVs,  FCVs only came into existence towards the end of the 20th century.  Residents of Northern California in particular  had no access to either FCVs or their hydrogen infrastructure until early in 2016.  Today Northern California has 8 operational stations, namely 6 in the SF Bay Area, one in West Sacramento, and one in Truckee 30 miles from Reno.  The Harris Ranch station at Coalinga links North and South California, an additional one at Santa Barbara is just reachable via US-101 from the SF South Bay, and there are ten more stations around the LA area.  In addition there are 20 more stations with committed funding in various stages of completion, 4 non-retail (bus) stations, 6 older stations soon to be acquired, and 15 stations targeted for the next round of California Energy Commission (CEC) funding.  The total at that point is 61 stations (not counting the 4 non-retail ones); and CEC has state authorization to fund an additional 39 stations beyond that for a total of 100 retail stations over the next five years.

An important benefit of FCVs is their refueling time of 5-10 minutes from completely empty to completely full.   Another benefit is that each additional $1000 in price gains about 6 more miles of range, double that of BEVs.  A third benefit is that, mile for mile of range, an FCV “engine” consisting of a full tank, a fuel cell, and a traction battery weighs about a third of a BEVs lithium-ion battery.  For example the Mirai’s drivetrain including a full tank weighs about 400 lbs while the Tesla Model S 85’s battery as packaged weighs about 1200 lb.

So what’s not to love about FCVs?

Well, if hydrogen could be made economically using electrolysis of water powered by solar electricity, FCVs would truly be ZEVs.

But hydrogen today is made far more economically by a process known as steam reforming of methane.  This process has two steps, the result of which is that a mole of methane (aka natural gas or NG) and two moles of steam (aka water vapor at high temperature) is converted to four moles of hydrogen and one mole of CO2.  In the sort of formula beloved of chemists this reaction amounts to CH4 + 2H2O → 4H2 + CO2.

(The only difference between a mole and a molecule is that a mole of any substance contains Avogadro’s number (about 6 followed by 23 zeros) of molecules of that substance.  Whereas a molecule of CO2 weighs only 44 times as much as a single proton, a mole of CO2 weighs 44 g (grams) or about an ounce and a half.  A mole of hydrogen atoms weighs 1 g while a mole of hydrogen molecules, H2, weighs 2 g.  And a mole of H2O weighs 18 g while a mole of CH4 weighs 16 g.  Hence in the reaction of the preceding paragraph, every 16 g of natural gas and 36 g of steam are converted to 8 g of hydrogen and 44 g of CO2.  Scaled up by a factor of a thousand (kilomoles in place of moles), and in the usual units for each chemical, about 800 standard cubic feet (0.8 mmBTU) of natural gas with steam from about ten gallons of water becomes 8 kg of hydrogen and about 100 lbs of CO2.

This analysis assumes 100% efficiency.  In practice steam reforming is 65-75% efficient.  If we optimistically take 75% as the efficiency then we only get 6 kg of hydrogen from each 800 scf of natural gas, but still yielding a hundred pounds of CO2.

That CO2 is the Achilles’ heel of the whole zero-emission premise of fuel cell vehicles!

Now the US Environmental Protection Agency (EPA) has measured the combined city/highway fuel consumption of the Toyota Mirai FCV at 67 MPGe (miles per gallon equivalent), which in the case of hydrogen is miles per kg of hydrogen. (A gallon-equivalent is a unit of energy equal to 121 megajoules (MJ).)  Production of 6 kg of hydrogen by steam reformation emits 100 lbs of CO2, whence driving the Mirai 100 miles entails an implied emission of (100/67)*(100/6) = 25 lbs of CO2 per hundred miles.

Each gallon of gasoline consumed by an ICEV is estimated to produce 18 lbs of CO2.  This is the implied emission from a Mirai driven 18/25 = 72 miles.  It follows that when using hydrogen 100% of which is produced by steam reforming, the Mirai’s implied emission is that of a car rated at about 72 miles per gallon.

Hydrogen vendors try to offset their carbon footprint by including some hydrogen derived from renewables produced without emitting CO2.  The vendor True Zero for example claims that 1/3 of its hydrogen comes from renewables.  With such a fuel the Mirai’s implied emission becomes that of an ICEV rated at about 110 miles per gallon.

For a 4100 lb car that can accelerate from 0 to 60 mph in 9 seconds and reach a top speed of 111 mph, 110 miles per gallon is remarkable.

The CO2 produced in this way is customarily emitted to the atmosphere.  If instead it is suitably sequestered by any of the methods described in the relevant Wikipedia article this method of producing hydrogen will further reduce the carbon footprint to that necessary for the power consumed by the method.

Another method of producing hydrogen is by electrolysis of water.  Two electrodes are immersed in slightly conducting water and a voltage over about 1.5 volts is applied across them.  This voltage is sufficient to break the electron bonds holding the hydrogen and oxygen atoms of H2O together.  The hydrogen and oxygen atoms thereby liberated then bubble up from respectively the cathode (the negative electrode) and the anode (positive).  The oxygen is discarded and the hydrogen produced in this way has an energy of about 60% of that used to create it and the oxygen.

When electrolysis is powered entirely by renewables such as wind or solar the resulting hydrogen is itself 100% renewable in the sense that no CO2 was emitted in its production.  If however it is powered by a typical electric utility, much of that electricity will have been produced by burning fossil fuels such as coal, oil, and natural gas.

The same holds of the electricity used to charge the battery of a BEV.  The crucial difference here is that whereas the Mirai gets only 67 MPGe, the typical BEV gets 50% more than that or around 100 MPGe.  Hence for electricity from fossil fuels an FCV like the Mirai has an implied emission 50% more than a BEV.

Increasing use of renewables in electricity narrows this gap.  In the limit when electricity production entails no emission of CO2, both FCVs and BEVs have the same implied emissions from this source, namely zero.

Posted in Uncategorized | 2 Comments

Joe Romm’s seven arguments against FCVs

Joe Romm has been arguing that fuel cell vehicles are a lost cause since even before he published his 2004 book The Hype about Hydrogen. On April 8, 2015 he published Tesla Trumps Toyota: The Seven Reasons Hydrogen Fuel Cell Cars Are Stalled. In that article he asked and then answered the following questions about “alternative fuel vehicles” (AFVs).

Consistently with his 2004 book, the general thrust of his answers strongly favors batteries over hydrogen as an alternative fuel. However his reasoning is based more on the same hypothetical arguments he has been using for decades than on any actual boots-on-the-ground experience with the latest technology in this sector. My own experience as a Toyota Mirai owner and driver has led me to quite a different conclusion, as I will explain after listing his questions.

1. High first cost for vehicle: Can the AFV be built at an affordable price for consumers? Can that affordable AFV be built profitably?

2. On-board fuel storage issues (i.e. limited range): Can enough alternative fuel be stored onboard to give the car the kind of range consumers expect — without compromising passenger or cargo space? Can the AFV be refueled fast enough to satisfy consumer expectations?

3. Safety and liability concerns: Is the alternative fuel safe, something typical users can easily handle with special training? [Presumably “with” should have been “without”. “Easily … with special training” is surely an oxymoron.]

4. High fueling cost (compared to gasoline): Is the alternative fuel’s cost (per mile) similar to (or cheaper than) gasoline? If not, how much more expensive is it to use?

5. Limited fuel stations (the chicken and egg problem): On the one hand, who will build and buy the AFVs in large quantity if a broad fueling infrastructure is not in place to service them? On the other, who will build that fueling infrastructure — taking the risk of a massive stranded investment — before a large quantity of AFVs are built and bought, that is, before these particular AFVs have been proven to be winners in the marketplace?

6. Improvements in the competition: If the AFV still needs years of improvement to be a viable car, are the competitors — including fuel-efficient gasoline cars — likely to improve as much or more during this time? In short, is it likely competitors will still be superior vehicles in 2020 or 2030?

7. Problems delivering cost-effective emissions reductions: Is the low-emission or emission-free version of the alternative fuel affordable? Are fueling stations for that version of the fuel affordable and practical?

You can see Dr. Romm’s answers at his article. Here are mine.

1. High first cost for vehicle. Romm broke this into a two-part question: (i) affordable, and (ii) profitable.

I can answer (i) very easily. The only BEVs available today that are more affordable than the 2017 Toyota Mirai’s monthly lease of $350 have a range about a third of the Mirai and take more than ten times as long to refuel even on a DC Fast Charger, let alone a slower Level 2 charger.

In a year or two Tesla may have a vehicle cheaper than a Mirai, namely the promised Model 3. But since Tesla today has no vehicle with the range of the Mirai at even twice the price, it would be a miracle if any battery electric vehicle manufacturer in the world, GM, Mercedes, Nissan, Tesla, whoever, could produce a 2018 BEV that beat the 2016 fuel cell Mirai on both price and range. More on this below.

As to profitability, what on earth is Romm thinking? In 1997 each $17,000 Prius cost Toyota an unprofitable $32,000 to make, yet ultimately the Prius was wildly successful and enormously profitable for Toyota. Romm himself has not only spoken highly of Toyota’s Prius but drives one himself.

Yet now we find him staking his credibility as an expert on hydrogen as a fuel on Toyota’s inability to perform that trick again. Why? Because of insider information that Toyota’s management has lost its marbles? No, this has nothing to do with Toyota. Romm has been convinced for more than a decade that no manufacturer in the world can make a fuel cell vehicle that will be competitive with battery electric vehicles judged by even one of his seven criteria.

Dr. Romm is a modern-day Lord Kelvin. In 1895 Kelvin stated emphatically that heavier-than-air flight was a physical impossibility even for a well-funded team of the best aeronautical engineers in the world, let alone two bicycle mechanics from rural Ohio. Scientific American bought lock stock and barrel the argument Kelvin and other distinguished scientists had been making, so much so that even a year after the Wright brothers historic flight at Kittyhawk they were still convinced it was a fraud. The French maintained the same conviction for five years until the Wright brothers eventually felt obliged to ship their plane to France and fly it under their very noses. Such is the inertia of belief based on theory even when confronted with evidence based on experience. More on this below.

2. On-board fuel storage issues (i.e. limited range). Again Romm breaks this issue into two questions, (i) range and (ii) refuelling time. For (i) the EPA range for the 2017 $350/mo Mirai is 312 miles. The only BEV in the world that exceeds this is the Tesla Model S P100D with an EPA range of 315 miles and an expected price of $134,000 when it eventually goes on sale—no lease has been set yet. The cheapest Tesla today is the S 60 at $685/mo with an EPA range of 210 miles, a mere 2/3 of the Mirai’s range!

For (ii) my Mirai refuels in 60 seconds per kilogram, so 5 minutes from completely empty to completely full, more typically 3-4 minutes. This is incredibly faster than any BEV. With a home charger the Tesla recharges overnight, while at a supercharger, even after 20 minutes or four times as long as the Mirai the Tesla is still only recharged to half its range, while a full charge take 75 minutes or 15 times as long as the Mirai!

3. Safety and liability concerns. There are three main concerns here, namely the environment, human toxicity with contact, ingestion or inhalation, and the hazard of fire.

Gasoline. Although aerosols used to be a problem with gasoline, today the environment is impacted primarily by oil spills from ocean-going drilling rigs and tankers, emission of the greenhouse gas CO2, and particulate matter (PM10 and PM2.5) from the exhaust, brake pads (thousands of tons annually), and tires. Hybrids with regenerative braking greatly reduce brake pad PM.

Regarding toxicity, in normal use gasoline itself does not have acute toxic effects. After combustion however vehicles emit a number of carcinogenic gases and fine particulate matter.

Regarding fire, whereas the typical autoignition temperature of paper is famously about 450 °F, gasoline’s autoignition temperature is a somewhat higher 500 °F or so.

Hydrogen. The tank-to-wheels environmental impact of hydrogen is limited to dribbling about half a cup of water a mile onto the road, and particulate matter from tires (relatively little from brake pads in models with regenerative braking). The impact of hydrogen production depends heavily on the production method and can range from negligible to high.

As to toxicity, hydrogen at room temperature is not strongly reactive and is therefore not acutely toxic to the skin, but, like helium, when inhaled may cause asphyxiation if it has displaced much oxygen. (CO2 is much worse: even at a mixing level of 6% by volume and displacing only nitrogen, so no risk of asphyxiation at all, it can still cause acute respiratory acidosis.)

Like gasoline, hydrogen is inflammable, with a relatively high autoignition temperature of 997 °F but a relatively low quantity of heat needed to trigger ignition. With these dangers of hydrogen in mind, especially the latter, states set stringent regulations on compressed hydrogen in both pumps and cars aimed at minimizing leakage and damage from violent decompression in the event of a component failure or collision.

Batteries. Like hydrogen, batteries have no environmental impact in normal use, but again like hydrogen their manufacture and charging can have various environmental impacts depending on the sources of their materials and electricity. Unlike hydrogen the disposition of spent batteries may have a further environmental impact.

With normal handling automobile lithium-ion batteries are no more toxic than those in laptops, cellphones, and cameras. Their much higher voltages however can be lethal if touched.

The higher energy density of lithium-ion batteries puts them at greater risk of thermal runaway. As a result of a number of fires on planes it is now standard to prohibit them as cargo on passenger planes. With such fires as a consideration, batteries used in electric vehicles are shrouded in fireproof packaging adding as much as a third to the weight of the constituent cells of the battery pack.

4. High fueling cost (compared to gasoline). As proponents of internal combustion engines will cheerfully tell you, gasoline is indeed cheaper than either batteries or hydrogen, at least with the present oil glut and cheap natural gas. However this analysis overlooks the costs of further atmospheric CO2. If those costs can be overlooked then why suffer the high cost and long recharge times of BEVs when gasoline vehicles are so much cheaper and easier to refuel? FCVs refuel as quickly as gasoline vehicles, are half the cost of a BEV with similar range, and suffer primarily from an infrastructure that in Northern California is only a few months old and is nonexistent in the 49 other US states. The rate of growth of that infrastructure is however well matched to the ability of FCV manufacturers to ramp up production.

5. Limited fuel stations (the chicken and egg problem). Chickens and eggs coevolved from amoeba via dinosaurs. Today fuel cell vehicles and their hydrogen stations are coevolving in a few regions such as Japan, Germany, and California starting from tiny quantities of both. The proper questions are, how fast will these new-age chickens and eggs grow in volume, and how will increasing volume lower costs? Anything that can double each year can grow a thousand-fold in a decade and a million-fold in two.

A year ago California had no publicly accessible hydrogen stations and hence no market for FCV manufacturers to sell into. Today California has 20 stations with at most one being down for maintenance at any given time, with the result that there is no shortage of hydrogen for FCV cars in California south of Fort Bragg other than for the few unfortunates showing up at the one that is down if any. (Toyota’s Entune app for iPhone and Android shows which stations are up any time you look.) At the moment of writing there is one station down, namely San Juan Capistrano, understandable because it has only been open for a few day and is still in its “soft opening” phase. Expect stations to open along paths towards the east coast as FCV manufacturers ramp up production, for example along I-10, I-40, I-70, I-80, and I-90. Relatively recent (2013) projections of growth and cost of the hydrogen highway can be seen at the 2013 National Renewable Energy Laboratory (NREL) report Hydrogen Station Cost Estimates estimating rollout and costs up to 2030 under various scenarios.

6. Improvements in the competition. Romm is quite right that FCVs have been hanging back well behind BEVs for years. But suddenly several FCV developers have rocketed past BEVs in the manner I described in 1-5 above. Perhaps this will encourage complacent BEV manufacturers to up their game in order to catch back up. Whether they can however remains to be seen, as this has now turned into a race between alternative energy technologies.

7. Problems delivering cost-effective emissions reductions. A fair question. I would argue that BEVs and FCVs are in a dead heat here, at least in California. According to PG&E, BEVs charged from their grid are getting electricity from 20% renewable sources. According to True Zero (my hydrogen vendor), FCVs refuelling from their stations are getting hydrogen from 30% renewable sources. This gap however is closed almost exactly by the fact that fuel cells are only 2/3 the efficiency of batteries: the EPA scores the Mirai at 67 MPGe and the Tesla S 60 at 100 MPGe.

BEV owners have the option of charging from home solar provided (a) they have home solar and (b) their BEV stays at home during the day. The latter however is a bit ridiculous. The home solar owner can argue that their system feeding the grid during the day is offsetting the carbon cost of charging the BEV by night. However everyone with home solar is doing the same thing, including those who own FCVs. Doing the math, an FCV owner’s home solar benefits the atmosphere no more or less than a BEV owner’s does.

This is not to say that Tesla’s cars have no advantages over the Mirai. In fact they have quite a few: more luxury, more interior space, faster (even ludicrous) acceleration, 20 mph higher top speed, and a less radical exterior design. What is odd about Romm’s criticisms of FCVs is that it completely overlooks all these pluses for the Tesla and instead finds fault in precisely those areas where the Mirai actually dominates every model of Tesla!

On the other hand there is one more very significant area stressed in Romm’s 2004 book that one might have expected him to ask about as an eighth question, phrased perhaps as “is the weight of the power train reasonable?” Although Tesla does not release that property of its battery packs, a member of the Tesla Motors Club purchased the 85 kWh battery from a salvaged Model S. Although the whole battery was too heavy for him to weigh in one piece, he was able to estimate its total weight based on weights of its major components, arriving at a figure of about 900 lbs for the cell assembly and 300 lbs for its protective packaging. This made the total weight 1200 lbs or 544 kg. The EPA range of the Model S 85 is 265 miles, making the energy density of the Tesla’s power train in terms of range 265/544 = 0.49 miles/kg.

The Mirai’s power train consists of two tanks weighing 87.5 kg, a 56 kg fuel cell, and a 36 kg 1.6 kWh nickel metal hydride (NiMH) drive battery very similar to that used in Toyota’s Camry Hybrid. Those plus 5 kg of hydrogen add up to 184 kg. The EPA range of the Mirai being 312 miles, the corresponding energy density for the Mirai comes to 312/184 = 1.70 miles/kg or 3.5 times that of the Model S 85!

Although Tesla has since replaced the rear-wheel drive S 85 with the all-wheel-drive S 90D, the former used to lease for $1,099/mo according to A Tesla with only 85% of the range of the Mirai and a power train three times as heavy therefore leased for more than twice the monthly cost of the Mirai. Today the S 90D as its closest replacement has brought the EPA range up to 294 miles, much closer to the Mirai’s 312, but its lease price of $1,104/mo has remained stubbornly close to twice that of the Mirai while its larger battery seems unlikely to have grown much lighter than the S 85’s.

To summarize, I have been unable to find any merit in even a single one of Dr. Romm’s seven arguments against FCVs.  But this then raises the question, how could he have got all this so wrong?

There is a simple answer. As my responses to Romm’s seven questions hopefully made clear, the fuel cell scene has changed dramatically in the 14 months since he asked those questions. Technology advances very quickly, making it essential for technology experts to keep abreast of new developments or risk losing their credibility.

Dr. Romm’s decades of valuable experience as an expert on hydrogen as a fuel is at risk of depreciating rapidly if not kept up to date.

Posted in Uncategorized | 2 Comments

Tony Seba’s six arguments against fuel cell vehicles

Tony Seba is a Stanford MBA with a BS in Computer Science from MIT currently residing in Northern California.  Quoting from his blog, “Tony Seba is the author of Clean Disruption of Energy and Transportation, Solar Trillions and Winners Take All, a serial Silicon Valley entrepreneur, and an instructor in Entrepreneurship, Disruption and Clean Energy at Stanford’s Continuing Studies Program. His work focuses on clean energy, entrepreneurship, market disruption, and the exponential technology trends, business model innovation, and product architecture innovations that are leading to the disruption of some the world’s major industries, such as energy, transportation, infrastructure, finance, and manufacturing.”

On January 15, 2015 Mr. Seba published a post on his blog listing six reasons why fuel cell vehicles (FCEVs) will be unable to compete with battery electric vehicles (BEVs) when they are introduced into Northern California.

A little over a year later several hydrogen stations were opened in the San Francisco Bay Area and Toyota began shipping its first few FCEVs, namely the Toyota Mirai.  This is therefore an appropriate time for a preliminary comparison of Seba’s six speculations with the reality of FCEV “tires on the road”.  Here they are, with the first two switched.

1) Electric Vehicles are at least three times more energy efficient than Hydrogen fuel cell vehicles.

2) Hydrogen is not an energy source.

3) You need to build a multi-trillion dollar hydrogen delivery infrastructure.

4) Hydrogen is Not Clean.

5) Hydrogen is not ‘Renewable’!

6) Hydrogen Fuel Cell Vehicles can’t compete with Electric Vehicles.

Here are my responses.

1) Electric Vehicles are at least three times more energy efficient than Hydrogen fuel cell vehicles.

Half right.  Seba’s factor of three in energy efficiency obtains for his “well-to-wheels” analysis.  When any such analysis is split up into “well-to-tank-to-wheels” it quickly becomes apparent that the huge variability of well-to-tank efficiency allows one to cherry-pick whatever values yields the desired ratio.

The advantage of tank-to-wheels efficiency is that the EPA has official numbers, namely the miles-per-gallon-equivalent, MPGe, for every vehicle.  A gallon-equivalent is defined as the energy obtained by combustion of one US gallon of gasoline, taken for definiteness to be 121.3 megajoules (MJ).

The various Model S Tesla BEVs have a very impressive fuel efficiency of 90-100 MPGe depending on the exact model.  The Toyota Mirai FCEV has a less impressive fuel efficiency of 67 MPGe.  In the tank-to-wheels analysis the Tesla therefore beats the Mirai in fuel efficiency, but only by a factor of 1.5, not 3.  What happens between the “well” and the “tank” is nowhere near as clear-cut and therefore not a sound basis for comparison.

2) Hydrogen is not an energy source.

Neither are rechargeable batteries.  How is this an argument against FCEVs?

3) You need to build a multi-trillion dollar hydrogen delivery infrastructure.

To accomplish what? In round numbers, a million FCEVs could be serviced by 10,000 stations costing $1M each, for a total of ten billion dollars. Hence if you replaced all 250 million registered vehicles in the US by FCEVs the 2.5 million stations needed would cost a total of 2.5 trillion dollars. At that volume it would be reasonable to expect the average FCEV to cost $20K, bringing the fleet cost to 5 trillion dollars or double the cost of the fueling stations.

But this won’t happen overnight. Instead both FCEV manufacturing capacity and hydrogen delivery capacity will each grow at rates that can only be speculated at today. California for example currently has 20 hydrogen stations, with $200M earmarked for development of many more.

Today’s oil-based fueling infrastructure for ICEVs (internal combustion engine vehicles) does indeed require a complex infrastructure for importing, refining, and transporting fuel around the country. The US does not need a massive infrastructure for importing and refining natural gas, and tra

BEVs and FCEVs have in common an important difference from ICEs, namely that they can be operated entirely off the grid when powered by solar PV.

Actually the opposite.  Unlike the centralized extraction and refinement of fossil fuels, hydrogen can be manufactured locally from natural gas (whose delivery infrastructure is already in place) or by electrolysis using electricity from either the grid (also already in place) or solar PV (no infrastructure needed at all).  Regions served by hydrogen plants can be as small as counties, cities, or even individual refueling stations.

4) Hydrogen is Not Clean.

The five currently operational True Zero hydrogen stations in the San Francisco Bay Area claim to be one-third renewable with plans to increase that.  Electricity from PG&E is only 20% renewable with plans to reach one-third by 2020.  BEVs recharged from the grid are therefore even less clean than FCEVs refueled at one of the five (soon to be seven) True Zero stations in the SF Bay Area.

5) Hydrogen is not ‘Renewable’!

How is this different from objection 4) ?

6) Hydrogen Fuel Cell Vehicles can’t compete with Electric Vehicles.

True enough for a Tesla in “ludicrous” mode.  A Mirai patrol car would look silly trying to catch a bank robber in a P85D.

But there are many competitions besides raw power, such as the capacity CAP, weight WGT, and energy density DEN of the “tank”, the range RNG after filling up, and the refuel time RFT.  For most buyers comparing BEVs and FCEVs, range and refuel time are likely to be the main and perhaps only concerns, far outranking energy efficiency in importance.  The units for these are as follows.

CAP: Tank capacity (MJ)
WGT: Tank weight (kg)
DEN: Density of storage (MJ/kg)
RNG: Range in miles (per EPA)
RFT: Refuel time in miles per minute
STN: Stations in the SF Bay Area

The following table compares the Tesla 85 and the Mirai for each of these.

…….. Tesla   Mirai
CAP     306     710
WGT     544     143.5-87.5
DEN     0.56     4.9-8.1
RNG     265     312
RFT     0.6-10  62.4
STN     4       5

CAP: The Mirai stores more than twice the energy of the Tesla’s battery pack.

WGT: The Mirai’s tank weighs 87.5 kg (all that just to hold a mere 5 kg of hydrogen!).  Adding in the fuel cell stack, which has no counterpart in a BEV, brings that up to 143.5 kg.

DEN:  Density (specifically gravimetric energy density) is simply the quotient CAP/WGT in MJ/kg.

RNG:  Contrary to what one might expect from the very different tank capacities, the Mirai only narrowly beats the Tesla in range.  This can be blamed mainly on the fuel stack, which introduces a substantial inefficiency absent from BEVs, which as noted above have no counterpart.

RFT:  By far the biggest win for FCEVs is the additional 62.4 miles of range obtained with an additional minute of refueling, which adds 1 kg of hydrogen to the tank (so to fill from empty takes 5 minutes).  A Tesla charging at home on 240 volts gains only 0.6 miles per minute.  On a supercharger it does much better, gaining about 10 miles per minute.

STN:  There are very few hydrogen stations in the San Francisco Bay Area.  True Zero has opened five and will open two more shortly.  Other vendors have committed to several more.

What often goes unmentioned in these discussions is that the Bay Area has even fewer Tesla supercharger stations, namely four, two on each side of the bay.  The SF Peninsula has stations at San Mateo and Mountain View while the East Bay has stations at Dublin and Fremont.

The more substantive advantages of the Tesla over the Mirai are a nationwide infrastructure of superchargers, much faster acceleration, and a cleaner dashboard with a much larger display.  Tesla drivers should be able to expand this list, my experience is limited to the Mirai.

Lastly there is the small matter of cost.  The MSRP of the 2016 models of the Tesla Model S and the Toyota Mirai are respectively $71,200 and $58,491.  Both can be expected to decrease with increasing volume. The “race to the bottom” will be interesting to watch over the next couple of years.

Posted in Uncategorized | 2 Comments

My new Toyota Mirai

On Thursday (5/19/16) I leased a Mirai from Toyota Sunnyvale (nearest Mirai dealer to my Palo Alto home).

Why Mirai?

1. ICEs emit many kinds of carcinogens into my neighborhood, see the book The Harmful Effects of Vehicle Exhaust described at Understandably this is not something either the auto industry or the oil industry is motivated to highlight as a beneficial feature of ICEs, and the consumer faced with the alternatives of public transport, a bicycle, or (gasp) a battery electric vehicle (BEV) is naturally in denial about it. Maybe hydrogen producers emit carcinogens, but not into my neighborhood AFAIK.

2. In 52 years of car ownership, including two new Mercedes, two second hand Mercedes, and a fleet of other vehicles from the UK, Germany, France, Sweden, Japan, and the US, it’s the first car I’ve ever owned that didn’t/won’t eventually drip oil.

3. Refueling time of 60 seconds per kg. That’s 4 minutes when your remaining range is 60 miles. It probably could refuel a fair bit faster in warm weather, this slow rate avoids the risk of icing up the nozzle in cold weather. When full (70 MPa = 700 atmospheres) it gives the remaining range as 342 miles, perhaps because it hadn’t yet figured out that I have a lead right foot. I’ve only had time for one refuel so far, I’ll keep an eye on it.

4. 300-340 mile range, adequate for me. I drive the 90 miles (in 90 minutes) from Stanford to the Stanford Hopkins Marine Station once a fortnight. Even though the nearest hydrogen stations to the latter are Campbell (66 miles to the north) and Coalinga (132 miles to the south, or 151 via I-5), there are currently two stations each within a mile of my normal route. While I wouldn’t normally do three round trips back to back, each of three hours (540 miles), I do have that option with this car—even with the longest-range Tesla I’d have to recharge at least twice on the way, however long that takes. For just the 90 miles, any BEV less than a Tesla may well have to recharge en route. And even with stations only just starting to come online in the last few months, already I could commute between Reno and Tijuana with stops for “gas” at just Sacramento, Coalinga, and LA. And if you want to take a 7-day vacation each year to places out of range of hydrogen stations Toyota kindly lends you a used Land Cruiser or whatever your trip needs (but you have to pay for the gas then, sigh).

5. Comfort of a loaded Lexus in a Camry form factor (but only four seats sadly). Everything in driver assistance short of keeping you in your lane if you fall asleep (it buzzes but that might not awaken you). Fully autonomous is still years away.

6. The price is right: including CA tax ($44/mo) plus optional GAP insurance and excess-wear-and-tear insurance ($47/mo), the driveaway payment after the $5k CA rebate is $500 (which I’d paid two months earlier as a deposit), the monthly is $499 + $44 + $47 = $590, and if after returning it (almost certainly) we get a different brand they ding us for $350 more. Fuel and service are free, but our insurer wants $1200/year for comprehensive coverage on top of that. Plus parking meters and toll roads.

7. Distinctive styling. I’m ok with it, and my Italian neighbors think it looks great compared to their conservative Highlander hybrid, which either reflects modern Italian styling tastes or they root for Darth Vader when they take their kids to see Star Wars movies. (Surprised George Lucas hasn’t already sued Toyota for infringing his stormtrooper design patent, but maybe he plans to get one—there’s a station near him in Mill Valley.) Anyway it’s only for three years, I’ll check out what Mercedes et al have on offer when the time comes.

8. Something about CO2 (insert belief here—infer mine from the 7.5 kW solar PV on my roof since 2008).

Posted in Uncategorized | Leave a comment

Hydrogen as a Fuel

I started this blog for the purpose of collecting and evaluating opinions about the future roles hydrogen might play in the global energy economy.

Basics first.

Energy density

Hydrogen is one of a number of forms in which energy can be stored. Practical alternatives to hydrogen include gasoline, propane, natural gas aka methane (in either liquid or compressed form), and batteries. Less practical alternatives include TNT, dynamite, and C-4, which can store a great deal of energy that is however hard to release at a usable rate.

The hydrogen atom consists of one proton and one electron. This makes it by far the lightest of all elements in the periodic table, so light in fact as to make comparisons of gravimetric (by-weight) energy density meaningless.  Hence I’ll stick to volumetric energy, in units of megajoules per liter (MJ/L).

To make hydrogen storage practical it needs to be compressed, typically to a pressure between 35 and 70 MPa (5,000 to 10,000 psi). The latter is standard for the so-called hydrogen highway and is therefore what I’ll assume here.

The first five entries in the following table are from Wikipedia

Coal 35 MJ/L
Gasoline 32.4 MJ/L
Propane 25.3 MJ/L
Wood 13 MJ/L
Hydrogen 5.6 MJ/L
Panasonic NCR18650B battery 2.56 MJ/L (6800 of these cells used in the Tesla Model S)
Tesla Model S battery 0.77 MJ/L (based on 85 kWh/(84″x48″x6″))
Tesla 400 kWh battery 0.22 MJ/L.
Tesla 10 kWh Powerwall 0.18 MJ/L
Posted in Uncategorized | Leave a comment