Childhood and fool’s gold
As children growing up in the 1950s in a rural area west of Denver, Colorado, my brother and I were fascinated by the Old West. A favorite adventure was to ride our bikes to the foot of South Table Mountain -- a mesa that lay between our house and the old mining town of Golden. We’d stash our bikes and hike around, examining rocks and playing at being old-time prospectors. There were washes running down the slopes from the top of the mesa. After summer storms, runoff was sometimes heavy enough to clear the washes and expose the rocks and gravel in their beds. We’d sort through newly exposed rocks and gravel, looking for gold!
We never found real gold, but on occasion, we found nuggets that looked like gold to us! The first time that happened, we threw the nuggets in our packs and took them home to proudly display the fruits of our day’s adventure. Our father, who had had a bit of actual prospecting experience as a young man in the 1930s, checked our finds. He examined our “gold nuggets”, made a show of scratching at them with a pocket knife, and finally informed us, “Sorry, boys, it’s not real gold, but these are nice samples of iron pyrite. Good find.” He then showed us various ways to tell the difference between iron pyrite crystals -- fool’s gold -- and real gold.
My brother and I were of course disappointed to learn that what we’d thought to be precious gold nuggets were a low-value common mineral. But Dad’s praise for our “nice samples” softened the blow. The lesson in mineral identification was a useful distraction. We took our finds, labelled them, and added them to our rock collection. The pyrite crystals were pretty, and we were young enough to enjoy pretending they were gold.
Fool’s green
These days, in the realm of clean energy, there’s something analogous to fool’s gold that we need to watch out for. I call it “fool’s green”: products and technologies that appear to be green, but on closer examination, aren’t. And unlike the case with fool’s gold, there’s no child’s play value in pretending they’re genuine. The stakes are too high.
Analogies are never perfect. Despite superficial similarities in appearance, the difference between gold and fool’s gold is unambiguous. “Greenness”, however, isn’t a clearcut either-or attribute of a product or technology. It comes in degrees. Solar PV panels deliver usable energy from sunlight, with no carbon emissions. That’s green, isn’t it? Yet solar panels are not without environmental cost. The production of polysilicon that underpins China’s dominance of solar cell manufacturing owes its low cost, in large part, to dirty coal-fired power plants. The plants are built nearby to provide China’s polysilicon factories with cheap, reliable, 24/7 power. Renewables can’t do that. Moreover, polysilicon production is only one link in a complex chain of resource extraction and carbon-emitting steps leading to installed solar panels. So are solar panels truly “green”? Or are they “fool’s green”?
Life cycle analysis has concluded that on balance, energy from modern solar PV panels does warrant a “green” label. Depending on location, it now only takes a few years of operation for panels to repay the resource and carbon debt of their genesis. But they’re not as green as one could wish. They’re best for applications like well pumping that have low capital costs and the flexibility to consume energy on an intermittent basis. Otherwise, the PV systems accrue further demerits on their “green-ness” scorecard. More resources must be spent for the backup power or energy storage capacity required to make them usable in an on-demand energy economy.
The green hydrogen conundrum
Electrolytic hydrogen presents a different set of issues. Whether the hydrogen is true green or fool’s green depends on the state of the power grid at the time the hydrogen was produced. Green hydrogen is usually thought of simply as hydrogen produced by electrolysis of water using renewable energy. But being produced using renewable energy isn’t enough. Until such time as the grid is fully decarbonized, energy used to produce truly green hydrogen must be surplus. There must be no competing load being served by a carbon-emitting energy resource that the energy used for hydrogen could have served instead. If there is, then the clean energy directed to hydrogen production will necessarily be offset by production from fossil resources that would not otherwise have been needed. The hydrogen isn’t green at all; it’s fool’s green.
The carbon footprint for fool's green hydrogen can be figured in various ways. There are many complications in the modeling when real considerations like transmission limits and energy storage capacity are factored in. Because of those complications, it’s arguable as to which way to calculate the carbon footprint is really the “right” way. But by any reasonable calculation, the figure is high. The thermodynamic minimum energy required to produce a kilogram of hydrogen by electrolysis at standard conditions is about 39 kWh. State-of-the-art PEM units are 80% efficient; that means 49 kWh of DC power per kg hydrogen. It would be a bit over 50 kWh AC power from the source of generation.
50 kWh is a lot of energy. Producing it from a gas combustion turbine at 40% thermal efficiency takes 450 MJ of thermal energy from natural gas. The combustion energy from natural gas is 50.6 MJ / kg, so we need 8.9 kg of natural gas per kg of hydrogen. That translates to 24.5 kg of emitted CO2. It’s fair to argue that only 60% of that should be charged to the production of hydrogen, since the hydrogen produced can return 40% of the electricity used to produce it if the hydrogen is used later to generate power. Hence the carbon footprint of electrolytic hydrogen produced from non-surplus renewable energy comes to 14.7 kg CO2 per kg H2.
It’s instructive to compare that 14.7 kg with the carbon footprint for hydrogen produced by other means. The overall steam methane reforming reaction (SMR) is:
CH4 + 2H2O + heat ⇒ CO2 + 4H2
From molecular weights, we see that it takes 16 grams of CH4 and 36 grams of steam to produce 8 grams of H2 and 44 grams of CO2. The mass ratio of CO2 to H2 is 5.5 to 1. That’s well under half of the carbon footprint of fool's green hydrogen -- before any effort to capture the CO2 output from the process. With CCS at a 90% capture level, the carbon footprint would be only 0.55 kg of CO2 per kg hydrogen.
Knowledgeable advocates of green hydrogen may at this point be crying “foul”. My figure of 16 grams of CH4 for 8 grams of H2 is based on chemical stoichiometry. It’s significantly lower than what’s typically reported for steam methane reforming. In addition, I only consider CO2 emissions proper, and ignore the CO2e from fugitive methane emissions. Very high estimates for the latter are the basis for Robert Howarth and Mark Z. Jacobson’s notorious assertion that using natural gas for heating or power generation is as bad as burning coal.
It’s true that figures typically cited for the amount of methane input to hydrogen output with SMR are a good deal higher than the 2 to 1 that I cited above. The SMR reaction is strongly endothermic -- which is to say it requires heat input to drive it forward. In the context of an oil refinery, where hydrogen is needed for desulfurization and hydro-cracking of crude oil, the most expedient way to supply that heat is by combustion of some of the natural gas feedstock. The SMR reaction is carried out inside catalyst-packed tubes inside a high temperature gas-fired furnace. It’s not the most efficient way to deliver thermal energy to drive the SMR reaction, but it’s cheap and easy -- especially when the cost of dumping a bit more CO2 into the atmosphere is ignored.
The figures that Howarth and Jacobson cite in their hatchet job on blue hydrogen translate to a 3.65 to 1 ratio of methane to hydrogen. That’s well above the minimum that would be needed if combustion of methane were used efficiently to drive the SMR reaction, but it’s not out of line with what’s used in cheap SMR reactors. It’s equivalent to 10 kg of CO2 per kg of hydrogen. Even that is less than the 14.7 kg CO2 for fool’s green hydrogen. However, it’s possible to do better.
In practice, nothing says that the enthalpy to drive the SMR reaction has to be supplied by combustion of methane in an inefficient furnace. In fact, when cheap renewable energy is available, there are strong advantages to supplying it by joule heating integrated into the catalyst-packed reaction tubes. An article in the 24 May 2019 issue of the AAAS journal Science details the advantages. The process is termed electrified methane reforming (EMR).
With EMR, no extra feed gas is burned to supply heat. Hence there is no flue gas, and no CO2 needing to be captured from it. The only CO2 is that produced in the reforming reaction itself. Virtually all of that is separated from the hydrogen output in the PSA step (pressure swing absorption) just upstream of the hydrogen output. Hence, in addition to generating less CO2 to begin with, EMR facilitates capture of the CO2 that is generated.
The danger of hype
The danger of all the hype about green hydrogen as the clean energy carrier for the future is that it will skew policy priorities in a counterproductive direction. Hype-driven policy makes it likely that much of what’s being invested for green hydrogen will actually go into production of fool’s green hydrogen. The money and materials invested won’t constitute merely an inefficient use of resources; manufacture and operation of these systems will actively increase net energy consumption and carbon emissions. It will not reduce them at all.
The reason that much of the electrolytic hydrogen produced under a “green hydrogen” banner will be fool’s green is straightforward. It’s a combination of economic factors conspiring with precedent. The economic factors are the need for competitive pricing of the hydrogen produced, along with the high capital cost of electrolysis equipment needed to produce it. The latter is a major factor in the cost of producing electrolytic hydrogen even when the equipment is able to be utilized at a high CF (capacity factor). It’s crippling if the equipment can only be utilized on occasions when renewable energy production would otherwise be curtailed. Yet just such restricted operation is required for true green hydrogen.
Given the strong economic incentive that exists to utilize electrolysis equipment at the highest practical CF, we should not be surprised to see electrolytic hydrogen producers doing just that. It’s a safe bet that the electrolysis capacity now ramping up for a green hydrogen boom will mostly be operated at close to 24/7. Most systems will draw power from the grid, irrespective of whether there is a momentary surplus of renewable energy. High utilization of equipment minimizes the cost of the hydrogen product. Minimizing that is necessary in order to bolster the claim that “green” hydrogen will soon be cheaper than hydrogen from fossil fuels. Belief in that claim is crucial for businesses to secure the subsidies they seek.
How it could play out
Undoubtedly electrolysis operations will be throttled back when actual energy shortfalls loom. Electrolysis is, after all, a discretionary load, and operation as a “virtual battery” lends credibility to the assertion that these facilities are valuable green energy assets. However, the shortfalls that will trigger throttling will be shortfalls in the overall energy supply, not shortfalls of clean energy alone. As a result, electrolysis will frequently be adding load to the grid at times when that added load can only be balanced by added supply from fossil fuels.
One might ask how electrolytic hydrogen producers could expect to get away with that type of operation and still label their product “green”. An easy answer: they could do it the same way that numerous companies now get away with claims to be “powered 100% by clean energy”. They’d purchase sufficient clean energy credits to cover the energy they consume in making hydrogen. Clean energy credits are largely an accounting gimmick employed for greenwashing. However, It’s a gimmick that’s become established. There’s precedent for using it.
Clean energy credits allow the purchaser to “call dibs” on a certain amount of the output from a participating clean energy provider. It doesn’t matter when or where the clean energy is actually produced -- beyond a requirement that the provider can’t sell credits beyond the amount of energy they’ll be supplying to the grid. It’s a neat way to sidestep the real difficulties of transmitting and storing intermittent power. Unfortunately, climate change doesn’t care about accounting gimmicks. It only cares about the radiative forcing from elevated atmospheric levels of CO2 and other greenhouse gasses. If we allow fool’s green hydrogen to pass as green, the result will be to increase net power consumption and global carbon emissions.
Keeping ourselves honest
It’s not my intent here to attack green hydrogen. Not if it’s truly green. There are ways to produce truly green hydrogen. But when we adopt policies to promote production of green hydrogen, they will have to be carefully formulated. If they don’t recognize and address the difference between green hydrogen and fool’s green hydrogen, it will be green hydrogen’s “evil twin” that profits. And it will be at the expense of climate change mitigation.
We might attempt to address the problem by requiring that, in order to qualify for subsidies, electrolytic hydrogen production facilities must be powered by dedicated clean energy resources. If systems are configured so that electrolyzers are physically unable to draw power from the grid, then their operation can never require increased generation from grid-connected fossil fuel resources. And, in fact, many of the green hydrogen projects now going forward do involve dedicated RE resources. But policies mandating that approach are not what we want.
One reason such policies are undesirable is that there will be periods when the production potential of grid-connected renewables will exceed available load. Even if that condition arises for only a few hours a week, we’d still like to be able to tap the excess capacity for green hydrogen. Otherwise it’s wasted.
Beyond that reason, there’s a more important consideration. Powering an electrolysis facility from dedicated resources isn’t enough to insure production of truly green hydrogen. Granted, isolation prevents power consumed from directly increasing fossil fuel use; it also allows the electrolysis system to be operated at the duty cycle of its dedicated resources. That will be less than it would be for full-time operation from the grid, but significantly more than for operation limited to periods when clean energy resources would otherwise be curtailed. However, it incurs an “opportunity cost” that hinders reduction in carbon emissions.
Clean energy resources that are dedicated to electrolysis are not available for replacing generation from fossil fuels. In terms of reducing carbon emissions, the latter will always be more efficient than production of hydrogen. So if connection to the grid is a viable option, it should be used. But then the green hydrogen facility is back to using only surplus energy.
The only way that I can see to robustly deal with the problem is to bite the bullet on power pricing. Wholesale power prices on the grid must reflect, in real time, the presence of any generation from fossil fuels. If fossil fuels are providing the marginal supply of electricity to the grid, then marginal load is being served 100% by fossil fuels. Its price should fully reflect that. The market would then force fool’s green hydrogen to carry a fool’s high price tag – just as it should.