Cement Rocks

The Question

What must be true for silicates to substantially globally displace carbonates as the feedstock for the current emissions-intensive cement manufacturing process?

The Problem/Opportunity

Silicates are wildly abundant, making up the majority of the earth’s crust. Locked in these minerals are resources that can be useful for a variety of industries. Silicates frequently contain elements like iron, aluminum, calcium, magnesium – elements that play a critical role in our carbonized, and potentially our decarbonized, supply chains. If we were able to harness these abundant silicates to make materials like cement, we could potentially displace emissions-intensive manufacturing.

The Hypothesis

A solution must have the following features:

1. It must be regulated as Ordinary Portland Cement (OPC)
2. The feedstock minerals must be globally abundant and accessible
3. The net cost to produce cement must be competitive with today’s process

A heuristic argument to support the hypothesis follows.

It Must be Regulated as Ordinary Portland Cement (OPC)

The global regulatory environment around the use of cement and concrete is highly varied, but generally there are two main types of codes that have been established.  Prescriptive and Performance-based.  However, even though a performance-based standard exists, there is perceived risk to shifting to new cement chemistries so without a major value-proposition there won’t be a shift from OPC until a major market boundary condition is changed (including regulatory prescription).  This fundamentally slows the potential adoption rate of novel chemistries, and for structural applications the timeline is even worse.

The Feedstock Minerals Must be Globally Abundant and Accessible

Cement is remarkably cheap.  At ~\$125/ton in the U.S., it is made locally because the cost of transporting it long distances exceeds the value of the product itself.  Current cement feedstock minerals are globally abundant and cheap to mine.  

The Net Cost to Produce Cement Must be Competitive With Today’s Process

In the absence of a regulatory shift that mandates the use of alternative cement manufacturing processes or chemistries, we will assume that the market will determine the dominant cement manufacturing process for a region based on (if not exclusively, then at least substantially) price.  In the event of a world where a price on carbon is applied to the cement industry, then the same logic holds modulo a consideration of the emissions footprint of various solutions.  

The Discussion

Where is All the Carbon?

Say you had some CO₂.  Like, totally cool, not a big deal, just, you know, maybe like you wanted to get rid of a little bit, stash it away, ‘cause you had a bit too much fun lighting things on fire for a while.  It’s really not a big deal, but say, hypothetically, you wanted to put it somewhere that it might stay put for, I don’t know, forever or something.  Where should it go?

Well, let’s start with a discussion of where it is today.  

If I were to ask you ‘hey, where is all the carbon on our planet?’, you probably would say ‘in plants, soils, animals, maybe some in the ocean… oh yeah, and way too much up there in the atmosphere’.  That’s true, there are some beautiful carbons in all of those places.  But it’s less than 0.003% of the carbon on the planet.  Where is all the rest? Alien civilizations in the core of the earth!? Nope.  Rocks.  It’s all about those rocks.  Specifically, a hell of a lot of carbonates^*Note, most of the carbon is probably in the core and lower mantle, but we’ll focus on the Earth’s crust and above.  

Where do these carbonates come from?  We can mostly thank marine organisms (along with non-biogenic precipitation) for our bountiful supply.  Over billions of years various natural processes have combined CO₂ with magnesium or calcium ions, in particular, to precipitate magnesium or calcium carbonate (MgCO₃ or CaCO₃).  Biological organisms are responsible for most CaCO₃ (you may recognize this mineral from the puka shell necklace you wore in high school when you thought surfing was cool that one summer).  When these organisms die, the carbonate sinks to the bottom of the ocean and eventually, through long geological processes, gets turned into a sedimentary rock.  The calcium ions mostly come from the chemical weathering of silicate rocks.^*CO₂, when dissolved in water, can form an acid, H₂CO₃, which can leach rocks  About 20% of the carbon in the crust is from buried organic matter.

The Grandest Carbon Cycle of All

The most potent and consistent carbon cycle of the planet is the silicate-carbonate cycle.  The chemical weathering of silicate rocks that enables marine organisms to form carbonates sequesters CO₂ over hundreds of thousands to millions of years, and through volcanism these carbonates are transformed (remember metamorphic rock from middle school?), eventually, back into silicates releasing CO₂ back into the atmosphere. Comprising a few hundreds of Mt of CO₂ per year, this flux of carbon may seem trivial in light of the Gt of flux that anthropogenic and non-anthropogenic emissions comprise annually.  However, over sufficiently long timescales (i.e. millions of years) the silicate-carbonate cycle is the most important feature of the natural global thermostat. 

What’s driving this cycle?  Why do silicates steadily get converted into carbonates?  

The answer, as is so frequently the case, is that they are thermodynamically favored to do so (if you ever don’t know an answer to a technical question related to physics, the answer almost always is ‘the thing will happen that lowers the energy state of the system’).  

Calcium is (abundantly) found in silicates as an oxide (CaO).  It turns out that calcium oxide (CaO) can react with CO₂ to form calcium carbonate (or calcite, CaCO₃) and that this reaction is favored (i.e. exothermic) at STP.  

CaO + CO₂ -> CaCO₃ dG = -131 kJ mol^-1

Why is There So Much CO₂?

The reason why CO₂ is something we like to make so much of is that it’s a low energy state for carbon.  This means when we take other forms of carbon, like coal (C), natural gas (CH₄), or ethanol (C₂H₆O) and turn them into CO₂, we get energy from the reaction which lets us do things like generate electricity, move cars around, or dull the pain of existence for a few hours.  The reason why there’s so damn much of it is that once carbon gets to CO₂, there’s almost nowhere for it to go that’s more energetically favorable.  Almost nowhere.  Actually, calcium and magnesium carbonates are more energetically favorable forms of carbon (at modest temperatures and pressures) than CO₂!  That’s why most carbon is in the form of carbonates in the first place! 

Wait, so if carbon wants to be in the form of a carbonate, and there is plenty of calcium and magnesium in the form of silicates out there, why is there so much CO₂ still in the atmosphere?  The answer comes down to kinetics, the rate at which the reactions described above occur.  Doing something is harder than doing nothing (one sympathizes).  Even if the reaction is ultimately energetically downhill, an activation barrier exists that prevents the reaction from occurring spontaneously without some activation energy.  Consequently, even though there are plenty of silicates out there, the chemical weathering process proceeds at a slow pace.  Increasing the concentration of CO₂ and increasing the temperature both speed up the rate of reaction.  This is why the silicate-carbonate cycle acts like a thermostat – the more CO₂, the more warming of the atmosphere, which results in faster formation of carbonates from silicates, which lowers the CO₂ concentration of the atmosphere, which has a net cooling effect.  

How Can We Turn CO₂ into CaCO₃?

I’ve long considered hijacking and accelerating the chemical weathering of silicate rocks to be a holy grail of Negative Emissions Technologies [see COMING SOON].  The thermodynamics are in our favor, all we have to do is get over those pesky kinetics.  Of course, we do this every day in countless industrial processes for a wide variety of reactions.  What are our normal tricks?  Well, we can increase the temperature, pressure, surface area, and concentration of the reactants, and we can try to find a catalyst that helps us take a shortcut through the activation barrier.  Unfortunately for us, these processes usually take energy, require capital (i.e. big chemical plants), and therefore cost money.  In a world where we aren’t yet ready to pay people to actively remove CO₂ from the atmosphere, there’s no profit motive to spend money to enhance chemical weathering processes.  

Damn.  If only we could use those carbonates for something useful so that people would be willing to pay for them.  

Well, it turns out that the most important component of the most widely-used manufactured good on the planet is calcium carbonate.  

Cement – a Great Place to Put Carbon

We use on the order of ~20-30 Gt of concrete every year as the foundational construction material for buildings and infrastructure.  Concrete is composed of cement (~9% by mass), coarse aggregate (~55%), fine aggregate (~28%), water (~8%), and air (though composition varies).  Cement, or, more specifically, ordinary portland cement (OPC), comprises clinker, supplementary cementitious materials, fillers, and additives.  The ‘clinker’ is the glue that gives cement its functional properties, and the ‘clinker factor’ is the percentage of the total mass of cement which comprises the clinker.  Clinker is made from the reaction of CaO with alumino-ferro-silicates to form the primary cementitious components of cement.  

When cement cures, calcium carbonate (along with important hydration products) forms and acts as glue to bind together the other components of concrete.  Through reaction with carbon dioxide from the atmosphere, approximately 20% of calcium atoms end up as carbonates over the lifetime of cement in a building, and if, during demolishing, the cement is crushed and post-processed the number can increase to almost 75%, though more typically it’s only an additional ~3%.  That means that today ~320 Mt of CO₂ are committed to being sequestered (over the lifetime of the cement structure) every year! (as a note, if we were to increase the lifetime carbonation rate of cement to, say, 75%,^*Certain cement applications, typically those requiring structural steel reinforcement, can be sensitive to carbonation rate thus limiting the degree (or depth) of carbonation of concrete desired in a given application through the use of an additive that increases the diffusivity of CO₂ in concrete, post-processing, or a change in chemistry we could increase the total annual committed sequestration to over a Gt of CO₂).  

Unfortunately, today we get our calcium for cement from no other place than carbonates themselves, mined directly from the earth.  We take CaCO₃ and turn it into CaO (the key ingredient in clinker) and CO₂.  This means that when we make cement, the first step is to release all of the CO₂ from the carbonates back into the atmosphere (in a process that also takes significant energy, which today we typically get from combusting fossil fuels and thus additionally contribute to CO₂ emissions), meaning that cement manufacturing is actually one of the worst offenders of greenhouse gas emissions.  Today, the 4.1 Gt of cement production comprises roughly 2.2 Gt of CO₂ emissions per year, or >4% of all anthropogenic emissions.  

Why do We Use Carbonates Instead of Silicates?

So why the heck do we use carbonates instead of silicates as our feedstock for the CaO that comprises the key ingredient of OPC?  

Cost.

It turns out that today it’s cheaper to get CaO from CaCO₃ than from naturally-abundant CaO-bearing silicates.  

Why is CaO from CaCO₃ cheaper than CaO from silicates?  

While there are minerals out there that have high concentrations of CaO (e.g. wollastonite), they are insufficiently abundant and geographically distributed (FAILS criterion 1).  The abundant calcium-bearing silicates (including basalts and gabbro) have, by mass, ratios of CaO varying from <5% to >30%.  Meanwhile, CaCO₃ has a mass-ratio of CaO of 56%.  That means you need to mind roughly 2 to 10 times more silicate to get to the same number of calcium atoms.

If we assume that mining costs are roughly constant per ton of material removed (a not unreasonable assumption for the most abundantly mineable rocks, but still probably not a perfect assumption), then that means the mining costs for the key input are roughly 2 to 10 times higher for silicates as a feedstock.  For the production of CaO today, the input and energy costs are roughly of the same order of magnitude, so assuming that all else is equal in the manufacturing process (e.g. capex) and even with reduced energy consumption per unit of CaO (a very generous assumption), the production of CaO from silicates is more expensive than from carbonates.  

See Black Box 1 for a high level model where you can control for various assumptions about the ratio of energy and mining costs for the two approaches.  

What if We Valued Byproducts?

The above analysis does not take into account the value or costs of byproducts associated with the CaO manufacturing process.  In the case of CaCO₃, the only byproduct is CO₂ (most commonly valued at \$0 today), which is simply vented to the atmosphere.  In the case of silicates, non-calcium bearing silicate minerals remain (we’re remaining agnostic to the process itself right now, just accounting for atoms).  If these residual minerals are not saleable, then they will need to be disposed of (most likely at a cost).  

How can we think about the influence of the cost or value of the process byproducts for the two routes to product CaO?  

For every ton of CaO produced by CaCO₃, there are .44T of CO₂ emissions from the direct process emissions.  There is an additional ~.4T of CO₂ emissions per ton of CaO produced from the energy used in the process.  

For every ton of CaO produced by silicates, there are between ~3 and ~18 tons of mineral byproducts.  Depending on whether this process requires energy, and the source of this energy, CO₂ emissions may be present due to the process.  But there are no direct emissions due to the chemistry of the originating mineral (silicates bear virtually no carbon).  

See Black Box 2 for a high level model that shows the consequences of the potential value (cost) associated with the byproducts for the two routes to produce CaO.^*In order to present estimates of the impact of the cost or value of byproducts, cost estimates for key inputs is required.  At \$0 cost/value, this model gives the same result as Black Box 1.

From these high level models we can begin to refine our statement of ‘what must be true for a new solution to matter’.  

If you’re not satisfied with high-level heuristic argumentation about the relative costs of producing cement from the silicate approach vs. the carbonate approach, I understand.  I, too, like self-flagellation.  To get a bit more refined in our analysis, we can leverage Black Box 3.  This dashboard lets you control a bit more of the key parameters that ultimately determine the economics of a cement manufacturing process.  Behind the scenes, high-level techno-economic and financial models for a standard cement plant and a modified plant (for the silicate->cement process) reflect the assumptions controlled by the dashboard.  This model reports cost on the basis of \$/ton of cement (not CaO) and estimates the emissions intensity of cement for the two processes.  

All else being equal (e.g. capex), according to the model a CO₂ price must exceed ~\$58/ton, or the mineral byproducts of a silicate process must be valued (on average) in excess of ~\$4/ton, for a silicate-CaO process to be cost competitive.  Of course, all else won’t be equal, but in the absence of a specific process to analyze, it’s as good an assumption as any.  It is worth comparing these values to the costs of alternative methods to decarbonize the cement industry (CCUS, for example).

Of course, as described above, cement, upon curing, absorbs CO₂ from the atmosphere.  Assuming the same product (OPC) this will be the same ratio for both carbonate and silicate processes.  However, it is worth noting that for silicate processes, so long as the energy inputs are limited in CO₂ emissions, there is a potential for a net CO₂ sequestering as part of the process!  In a world with a price on carbon, this could create a higher value for the cement, but since the same amount of CO₂ is sequestered as is captured by the standard process (thus reducing its net CO₂ footprint), the above relative analysis still holds.  

Of course, if the required energy inputs for silicate valorization are NOT zero-carbon, then any increase in total energy requirements for processing of silicates may detract from the carbon story.

How Big are the Potential Consequences of a Shift?

Per our prior calculation, if all cement were to be generated by extraction of CaO from silicates, and we assume no CO₂ emissions associated with the production of CaO from this process, the sequestration potential would be between ~0.3 and ~1.1 Gt, or ~0.6-2.2% of anthropogenic greenhouse gas emissions!  What’s more, if we were willing to pour additional concrete simply to store CO₂ (meaning we would be using this as a means for a negative emissions technology) the price per ton of CO₂ sequestered would be ~\$260-\$860 (assuming no revenue from byproducts, zero-carbon electricity, and dependent on rate of carbonation).  This is an effectively infinitely-scalable process that can sequester carbon for millions of years in a thermodynamically stable way.  

Of course, if we were simply interested in optimizing for (effectively) permanently sequestered carbon per \$ of spend, we might look instead to a similar process involving olivines as an input rock (which can be nearly 50% MgO by mass).   

See [COMING SOON] for a discussion of the other paths for sequestration.  NONE of the other truly scalable solutions are thermodynamically stable (and hence secure against potential re-emission) in the way that carbonate storage can be.  

As a back of the envelope – just how much rock would we have to mine to offset the ~10 Gt/yr that we think we’ll need to sequester with negative emissions technologies?  Assuming the mass ratio of CaO in rocks to be 10%, and assuming that 100% of the CaO is carbonated, we would need on the order of 127 Gt/yr of mined rock.  This is only ~2.7 times the amount of currently mined rock by the aggregates industry today.  This is equivalent to mining ¾ of a Mount Everest every year.  Of course, if we wanted to use this technology for simply negative emissions, we might consider other options to accelerate weathering (like mining, grinding, and depositing calcium-bearing silicates into the ocean where they will form hydroxides that will react with the CO₂ in the water to form carbonates with a nice side-benefit of reducing the pH at the same time).  If we were to do the same scheme with magnesium-rich olivines, for example, we would only need to mine on the order of ~0.5 times the amount of currently mined rock by the aggregates industry today.  For more on this and other crazy schemes, see [COMING SOON].  

Other Vectors of Valorization 

Silicates can be valorized for other industries beyond cement. The iron, aluminum, magnesium, potassium, phosphorous, etc. content of these rocks may enable additional value streams. Consequently, should silicates be leveraged as a mined resource, the business model may look more like a ‘rock refinery’ than like a cement plant.

The Conclusion

Silicates are an abundant source of CaO, the principal ingredient of cement.  However, the concentration of calcium in these rocks is less than than the conventional source of CaO, calcium carbonate.  Consequently, more rock needs to be mined and processed per ton of produced cement if silicates are used as the source of CaO.  But advantages exist as well –   silicates bear minimal carbon, so there are no process emissions and the minimum thermodynamic energy required to produce CaO from silicates is lower than that of carbonates.  The cement produced by the silicate route is chemically and functionally identical to the cement produced by the conventional route.

If the byproducts of cement-making processes (CO₂ in the conventional case and alumino-ferro-silicates in the case of silicates) are valued, then using silicates as a feedstock for cement manufacture may be cost-advantaged. 

The abundant magnesium and calcium oxides found in silicates can react with atmospheric CO₂ to form carbonates, which stably sequester the CO₂.  Consequently, using silicates in lieu of carbonates as a feedstock for cement provides a potential path to zero- and even negative-emissions cement production.

References