Cement is the gray glue that binds our cities and industries together. It makes tall buildings possible and foundations strong. It helps us bridge rivers and valleys and keeps wind turbines upright. We don’t have a replacement for it. It’s not going anywhere.
But it’s also one of the largest sources of CO2 emissions globally, with estimates ranging from 5% to 12% depending on the source. The common solutions today for emissions are just like putting a little hydrogen in a natural gas line, an inadequate response to the climate crisis.
So what can we do about it? Do we have actual solutions? What will they cost? When will they be in place?
The greenhouse gas emissions from cement come from a couple of major things and a few minor things. Let’s deal with the minor ones first, because they have obvious solutions.
Cement requires quicklime, which is manufactured from limestone (the hard bit to be covered next). Limestone is heavy rock which has to be mined and shipped, both of which use fossil fuels right now. And then the heavy cement has to be trucked to its point of use, using more fossil fuels. Electrification of mining equipment and transportation is under way, as is decarbonization of the grid. That part is just a matter of time. It’s only about 10% of the emissions challenge.
However, the process of making cement isn’t a fully solved problem right now, specifically the manufacturing of quicklime from limestone. For context, quicklime is 66% to 90% of concrete, according to varying sources, so this is the majority of emissions. Manufacturing quicklime involves heating up the limestone in a kiln to hundreds of degrees in order to bake off the excess carbon and oxygen leaving only the quicklime. And those are the two problems, the source of heat and the CO2 from the limestone. Energy represents about 40% of the emissions challenge, and the CO2 baking off of the limestone is the other 50%.
Emissions From Making Heat
First, the heat comes from fossil fuels and second the carbon and oxygen that bake off combine into CO2. For an assessment of the use of concentrating solar power for this purpose (a purported Gates’ investment use case and a failed idea), I wrote this:
“Lime kilns for large scale production of quicklime are rotary drums up to 13.5 ft or 4 meters in diameter and 400 feet or 122 meters long. Ignition of natural gas occurs inside the body of the rotating drum with high heat flames roughly 3 times the length of the diameter of the interior of the kiln, so in the case of the bigger kilns, that’s a jet of flame roughly 40 feet or 12 meters in length and about half the diameter of the kiln in width.”
Yeah, natural gas inside a 100+ meter long rotating drum. There are vertical kilns and diagonal ones. The heat isn’t actually a problem. We can make as much heat as we want in a dozen ways with electricity. Here’s a patent for an electric-heat lime kiln for example. Apparently a Newfoundland cement plant, North Star Cement, was electrically powered, although it was shut down a few years ago.
This part is a solved problem. All it is is more expensive than using dirt cheap fossil fuels. Electricity is already starting to drop in price in some places with more renewables on the grid because they are so cheap to operate. Projections are for the wholesale cost of electricity to be in the $20 per MWh in much of the world by 2050. But right now, natural gas is really cheap in many places. Its distillates are much higher margin, but often companies are basically giving away the gas for the cost of distribution plus a bit of markup. It’s a bulk commodity revenue play and very cheap for the BTUs it outputs. Carbon pricing will be tipping the balance over the coming years, with electricity getting cheaper and gas getting more expensive, then the cement market will be buying electric kilns instead of gas-fired ones.
Let’s take a ton of cement. Natural gas has been fluctuating between $2 and $5 per million BTUs. Modern lime kilns take 6–8 million BTUs per ton, so that’s $12 -$40 per ton energy costs.
BTUs can be easily converted to MWh, and 6–8 million BTUs turns into 1.8 to 2.3 MWh. When you start getting into $20 per MWh electricity, that means $36 to $46 per ton energy costs, which is higher. But then you start pricing carbon, the economics change pretty quickly.
Currently, Canada has a really low carbon price. It turns into $1.9864 CAD per gigajoule ($1.50 USD) at the currently $20 CAD per ton of CO2e. Those BTUs equate to 6.3 to 8.4 gigajoules, so that turns into another $9.40 to $12.60 USD per ton by itself. The cost of energy for a ton of cement using natural gas has suddenly gone to $21.40 to $52.60 USD, which overlaps with electricity.
But Canada’s carbon price is going up to $30 CAD then $40 then $50 over the next few years, so the related cost adder for a gigajoule of natural gas is increasing as well. That means that in 2020, quicklime heating energy from natural gas will be $26- $59 per ton. Then in 2021 it will be $31-$65 per ton. And at $50, it’s about $36-$71 per ton, so it’s usually cheaper to use electricity if it gets down to $20 per MWh electricity.
But we don’t have $20 per MWh electricity yet. Industrial rates are often in the $40 per MWh range. That’s still $72 – $92 per ton.
This is one of the reasons why carbon pricing is inadequate by itself. Very, very few jurisdictions have ever managed to get carbon price up to $30 CAD per ton, and Canada just had one of the federal elections where that was an issue, and the people who wanted to axe the carbon price had a good shot at winning but thankfully failed. Australia wasn’t so lucky. They got to $23 AUD per ton a few years ago and then the sensible government was ditched by the conservative one which had promised to axe the carbon tax, the more typical pattern.
You can certainly see that by 2050, sensible carbon pricing and lots of cheap renewable electricity on the grid will mean that the economics of heating with electricity will be cheaper than the economics of using natural gas, but we aren’t there yet. And to be clear, this means that each ton of quicklime used in cement will be more expensive.
So yes, this is a solved problem technically, but not solved problem economically or politically today.
Emissions From CO2 From The limestone
This problem isn’t going away easily. There are a lot of efforts to make lower CO2 cement, but mostly it’s changing the proportions of other fillers that are lower CO2, mostly leftover fly ash from burning coal, which we have in huge toxic piles dotting the world.
However, we still need quicklime, and the only readily available source is limestone. And as stated, the carbon and oxygen that bake off are in the form of CO2.
However, we have the technology for that as well. Capturing CO2 at the source of emissions works, and we know how to pump it underground. The problem is once again not technology, but expense. The Canadian Boundary Dam coal plant in Saskatchewan, Canada is instructive. Basically, the cheapest that they could make it even with enhanced oil recovery revenue turned the electricity into $145 per MWh, about doubling it.
The per-ton of CO2 cost for at source capture and real sequestration is $100 – $140 per ton of CO2. How much CO2 does a ton of limestone emit? About half a ton per the IPCC.
So every ton of quicklime comes with half a ton of CO2, which costs about $50-$70 to capture and sequester. Assuming the maximum Canadian price on carbon and 100% carbon capture and sequestration costs for CO2, that brings the full cost per ton of quicklime up to $86-$141 or 4-6 times more expensive than it is today.
Changing Cement & Using Less Of It Are Options
That ton of cement is looking more and more expensive, isn’t it? If carbon were priced, a lot less cement would be used.
There are other aspects as we start looking at alternatives to quicklime for cement. Some don’t set at room temperature, but must be heated to 50 to 90 degrees Celsius, making them untenable for the vast majority of construction. Imagine heating a large bridge across a river for example. Others only lower total emissions by 10%, so not a particularly useful solution. Others require a lot of aluminum and have roughly 25% of total CO2 emissions, but the aluminum makes them expensive as well. It’s been fluctuating around $1,900 per ton over the past year, so the cost can be challenging. A cement formulation that absorbs over its life as much CO2 as is emitted during manufacturing uses magnesite, but the global market and supply for magnesite is a tiny fraction of cement, so it’s incredibly hard to scale up. Another process out of Germany cuts both heat and quicklime substantially, leading to a 50% reduction in overall CO2 emissions, but it’s a much more complex process that has never made it out of the lab and a production level greater than 100 kg per day.
Geopolymer approaches, which include blending things such as fly ash from coal generation, appear to be the most effective from a performance and cost perspective, and do provide a useful pathway for getting rid of the fly ash itself over the next 30 years, but of course we’ll be stopping coal generation and new supplies won’t be being created.
And at least in Canada, one major user of cement is going to be diminishing. Buildings up to 12 stories are safely built using engineered hardwood, substantially reducing the cement requirements. BC has already approved it and Canada is making it national next year.
Regardless of anything else, the primary glue that holds our buildings, bridges, and infrastructure together is going to become more expensive in a low-carbon world. This has fairly significant implications for the cost to build most of the infrastructure we depend on.
Note: I’ve reached out to experts working closely in the field of embodied carbon for an opinion on the most cost effective solution for this that they are aware of. Should they respond, I’ll update the article.