So, have you heard the story of the boy who cried “fuel cell vehicle feasibility”? Don’t worry, this isn’t a re-telling. It’s an investigation of one of
the hurdles fuel cells face in their quest for respect from the
cleantech crowd. Namely, can the hydrogen ever come clean?
Most hydrogen produced today comes from fossil natural gas through
the steam-methane reforming, or SMR, process – and unless hydrogen can
be produced competitively from renewable energy, fuel cell technology
risks hitting fossil status, itself.
Figure 1. SMR and Electrolysis reactions for hydrogen production
Disclosures
Fuel cells are a polarizing technology (hey, like sunglasses!) thanks
in part to boys who cried “fuel cell vehicle feasibility” decades too
early. Bullsh*t claims on both sides
don’t help the situation, and Toyota’s astonishingly ill-advised
anti-EV ad campaign has cost it some of the goodwill it won through the
Prius.
I have worked on fuel cells for the past 15 years, and do see promise
in the technology, which has proven to be a good fit in certain niches.
For example, over 100,000 micro-CHP (combined heat and power) fuel
cells have been sold in Japan. These provide all of a household’s hot
water needs, along with its first 700 watts or so of electricity. While
all-solar households like my in-laws’
may become the norm in Japan, micro-CHP holds promise for multi-unit
dwellings (condos/apartments) and in colder, cloudier climes such as
Europe’s.
I also own a plug-in hybrid electric vehicle, chronicle Canadian EV sales, and see batteries as our best, fastest way to take an enormous bite out of worldwide transportation emissions. Now, despite their higher costs and lower “wind turbine-to-tire”
efficiency, I still believes fuel cell vehicles will play a supplemental
role in decarbonizing transportation. In many respects, my opinion aligns with that of Greenpeace, which in section 11.3.1 (page 288) of its Energy [R]evolutions 2012 report noted that:
“In the future it may not be possible
to power LDVs [light duty vehicles] for all purposes by rechargeable
batteries only. Therefore, hydrogen is required as a renewable fuel
especially for larger LDVs including light commercial vehicles.”
The Union of Concerned Scientists largely concurs.
Why Worry Where The Hydrogen Comes From?
Since fuel cell vehicles will be lucky to achieve even rounding-error
levels of global vehicle sales in the coming decade, readers would be
forgiven for thinking the question of where the hydrogen comes from, is
moot. Why talk about tomorrow’s hypotheticals, when we can tackle
today’s troubles?
Running the numbers, a few tomorrows from now – the end of 2017 – Toyota plans to have sold up to 3,000 Mirai FCEV’s in the United States. The car gets about 60 miles per kg hydrogen (H2),
so assuming drivers cover 12,000 miles per year (a bit less than the
average Californian, to reflect the limited fueling infrastructure), we
can estimate Toyota’s fuel cell forerunners would use about 1.5 tonnes
of hydrogen per day. [1] In two years. Maybe. If everything goes to
plan. Which it often hasn’t, in the past. And even the present. (I love my fuel cells, but I need to be honest, eh?) But there is a rapidly growing market where fuel cells already use 4 times as much hydrogen today – more than 10,000 passenger FCEVs’ worth. It’s the warehouse forklift market.
The Forklift Factor
Segment leader Plug Power had about 5,000 fuel cell forklifts in
distribution centers around North America at the end of 2014, which
it expects will grow to about 8,000 at the end of the year.
[Disclosure: your contributor owns no stock in Plug Power,
Toyota, or any other company mentioned in the article. Besides which,
investing is one of his core incompetencies!]
However cheap lithium-ion batteries get, fuel cells are likely to
dominate this niche because tearing the battery-charging areas out of
their warehouses effectively allows companies to “enlarge” each one by 5
to 10%. The resulting productivity savings are bigger than the cost
premium of fuel cells and hydrogen, over lead acid batteries and
electricity.
But back to the hydrogen. If generated from renewable electricity, it would probably take a 40
to 45 MW wind farm to electrolyze those 6 (and rising) daily tonnes of
hydrogen. [2] While that’s not huge, it’s not tiny, either. Add a few more years of
forklift growth and throw in some FCEVs, and we’re looking at enough
possible electric demand that hydrogen producers (specialty gas
companies such as Praxair or Air Liquide) could start signing
PPAs (Power Purchase Agreements) with renewable energy developers. Of course, that will only happen if renewable hydrogen from
electrolysis proves cheaper than fossil hydrogen from natural gas. So,
how do these stack up?
Fossil North America …
The cost to produce chemicals in large quantities generally
simplifies to the feedstock costs plus the amortization of the capital
costs incurred to build the facility. (Labour costs tend to approach
0%.) As for capital costs, Table 6 of the Nicholas/Ogden paper referenced
in endnote [2] suggests that the capital costs of new electrolyzers and
new SMR units are broadly comparable, with electrolyzers perhaps being a
bit higher.
Which leaves feedstock costs. Building on Table 7 of the paper, I have estimated the feedstock
costs for renewable and fossil hydrogen. The screengrab below comes from
a spreadsheet available at www.tinyurl.com/FCStats (on the “Renewable Hydrogen” tab). Given North America’s fracking-assisted low natural gas prices,
fossil hydrogen will be cheaper than renewable hydrogen for the
foreseeable future. A hydrogen producer would have to sign a wind energy
PPA for the equivalent of 1.1 cents/kWh to have a shot.
Unfortunately, PPAs in the United States are still generally north of 2 or 3 cents – and that’s with a small assist from the Wind Production Tax Credit. Carbon taxes are unlikely to make a difference, as a $20/tonne CO2
price would only increase the cost of fossil hydrogen production from
about $0.63 to $0.74 cents per kg. For renewable hydrogen to be
competitive at 2 cents/kWh electricity, the carbon price would have to
be north of $100/tonne. (We can dream…)
An October 2010 white paper from Praxair [3] does estimate that
50–90% of the CO2 produced in SMR plants could be captured at a carbon
price of $40 to $50/ton. Unfortunately, hydrogen plants don’t generate
enough CO2 to make transporting it worthwhile, and it’s entirely
possible that the gas would get injected into aging wells … where it
would help push more oil out of the ground (a process known as “enhanced
oil recovery”).
So, absent far higher natural gas prices (whether from drilling
economics or carbon pricing) it’s likely that in the United States,
hydrogen will continue to be fossil-derived. Perhaps this is why the California Energy Commission legislated that 33% of the hydrogen sold
through stations it subsidized had to be derived from renewable
sources; policymakers may have realized that without the legislation,
it’d just be another fossil fuel.
Figure 2. Hydrogen feedstock costs; full spreadsheet (and links) available at www.tinyurl.com/FCStats
… and Renewable Europe?
Europe, however, is a different story. Natural gas costs are roughly
triple – and countries are keenly interested in being less reliant on
Russia for supply. If European specialty gas plants could source renewable electricity
for about 3 cents/kWh, when it comes time to increase production or
replace aging facilities, it might just be cheaper for them to choose
electrolysis over SMR.
Wholesale electricity prices in Europe aren’t quite that low, but
renewable electricity certificates in Sweden and Norway are already in
the right range, with upcoming wind projects expected to drive prices
lower. Wrinkles do remain – would utilities allow industry to access
wholesale rates or renewable electricity certificate pricing? – but
costs are in the right ballpark. Perhaps progressive utilities with
access to cheap wholesale electricity might partner with electrolyzer
and/or specialty gas companies with the right technology and
infrastructure. Or maybe carbon prices and/or low-carbon fuel standards
could also tip the balance towards renewable hydrogen when new
facilities are planned.
Wrinkles notwithstanding, it does seem that, in parts of Europe,
renewables-derived hydrogen may become more compelling than fossil
hydrogen –not just on an idealistic or moral basis, but on a mercenary,
profit basis as well. Whenever and wherever that crossover happens, local fuel cell
proponents will surely sigh with relief that they no longer have to
acknowledge valid criticisms about where their fuel comes from. (Toyota
will no doubt bring out a new ad in its “Fueled by Everything” campaign
to commemorate the event.)
And while carbon pricing may get credit for helping hydrogen “come
clean,” especially if attempts are made to account for fugitive methane
emissions, it would be more accurate to give the recognition to the wind
industry, for steadily driving costs down year after year as it has marched down the experience curve.
After all, without those cost reductions leading to the aggressive
buildout of wind capacity, which in turn pancaked wholesale electricity
prices, few people would consider making hydrogen from cleanly
electrocuted water, and it would ultimately continue to come from dirty
holes, dug deep in the ground.
Acknowledgements: the author would like to thank the following
individuals for their assistance as the article evolved: Al Burgunder
(Praxair); Andy Marsh (Plug Power); David Reichmuth (Union of Concerned
Scientists); Geoff Budd (ITM Power); Michael Nicholas (UC Davis). All
errors and inaccuracies are the author’s.
Endnotes:
[1] 3000 cars x 12,000 miles/year x 1 kg H2/60 miles = 600,000 kg H2/year = 600 tonnes/year
600 tonnes/year x 1 year/365 days = 1.6 tonnes/day
The 60 mile per kg H2 figure comes from the Mirai getting 300 miles per 5 kg fill-up, in Toyota’s press materials.
[2] It takes about 55 kWh of electricity to electrolyze and compress 1 kg H2, according to the following paper:
Michael Nicholas and Joan Ogden, An Analysis of Near-Term Hydrogen Vehicle Rollout Scenarios for Southern California, UC Davis Institute for Transportation Studies, February 2010.
To generate 6000 kg H2 per day, then, would require:
6000 kg H2 x 55 kWh/kg H2 = 330,000 kWh = 330 MWh
There are 24 hours in a day, so the hydrogen would require a continuous power of:
330 MWh/24h = 14 MW
Assuming a wind farm has 33% capacity factor, to provide an average 14 MW per day its size would have to be:
14 MW/0.33 = 42 MW
Given the uncertainties, call it a 40 to 45 MW wind farm.
Going from compressed to liquefied hydrogen would take more energy,
and a bigger wind farm. But that energy is also required to liquefy
hydrogen from Steam Methane Reforming plants.
[3] Dante Bonaquist, Analysis of CO2 Emissions, Reductions, and Capture for Large-Scale Hydrogen Production Plants, Praxair, Oct 2010.
http://cleantechnica.com/2015/07/04/wind-energy-back-door-for-fuel-cell-respectability/