Mitigating CO2 emissions of concrete manufacturing through CO2-enabled binder reduction

Past studies on CO2 utilization in the concrete industry have primarily focused on maximizing sequestered CO2, while focusing less on CO2 avoidance possible by reducing binder use through the addition of CO2 to concrete formulations. In this paper, we study the net CO2 reduction and cost benefits achievable by reducing binder loading while adding CO2 via three approaches: carbonation during curing, carbonation during mixing, or carbonation with recycled concrete aggregate. These techniques are evaluated for a cohort of concrete formulations representing the diverse mixture designs found in the US ready-mixed and precast industries. Each formulation is optimized for reduced binder loading where the use of CO2 directly in the formulation recovers the lost compressive strength from reduced binder. We show that over an order of magnitude more CO2 can be avoided when binder reduction is jointly implemented with CO2 utilization compared to utilizing CO2 alone. As a result, nearly 40% of the annual CO2 emissions from the US concrete industry could, in principle, be eliminated without relying on novel supplemental materials, alternative binder, or carbon capture and sequestration. The recently amended 45Q tax credit will not incentivize this strategy, as it only considers carbon sequestration. However, we find that the saved material cost from reduced binder use on its own may provide a significant economic incentive to promote the joint strategy in practice. We conclude that the real value of CO2 utilization in concrete hinges on exploiting CO2-induced property changes to yield additional emission reduction, not by maximizing absorbed CO2.


Introduction
A recent report from the Intergovernmental Panel on Climate Change regarding climate-related impacts of global warming of 1.5°C above pre-industrial levels, found that net global CO 2 emissions need to drop below zero starting mid-century [1]. The concrete industry is one of the major obstacles to achieving a net negative economy; with a current CO 2 reduction target limited to only reaching 25% below current levels by 2050 [2]. Today, global annual concrete consumption stands at 10 billion m 3 , or 25 billion tonnes, which makes it the most utilized engineered material around the world [3,4]. Despite the need to reduce global CO 2 emissions, concrete consumption is expected to increase by 12%-23% by 2050 compared to 2014 [2].
Reducing CO 2 emissions from concrete is challenged by the CO 2 emitted during the production of ordinary Portland cement (OPC), a main binder of concrete. Cement production alone represents about 85% of the CO 2 emissions from concrete manufacturing [3] and contributes around 7% of global CO 2 emissions [2,5] today. Approximately 60%-70% of this CO 2 is generated during limestone decomposition, also known as the calcination process, and thus cannot be mitigated by transitioning to a low-carbon energy supply chain. Therefore, efforts to curb CO 2 emissions from concrete fabrication have focused on substituting clinker-rich OPC with supplementary cementitious materials (SCMs), which are usually Original content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence.
Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
repurposed industrial byproducts providing similar mechanical properties to OPC. There is widespread agreement in the literature that incorporation of SCM offers one of the largest CO 2 mitigation opportunities until other options, such as carbon capture and sequestration (CCS) or alternative binders, become commercially viable [2,3,[6][7][8][9][10][11][12][13][14][15]. However, expanded use of SCM is unlikely to meet long-term CO 2 reduction goals in the concrete industry. The most common SCMs include fly ash (FA) and granulated ground blast furnace slag (GGBS), which are already being utilized near to their maximum potentials [16]. In the US, the utilization of GGBS has been supply-limited for decades [16,17]. The utilization rate of FA in US concrete reached 64% in 2017 and can potentially increase in the short-term due to growing demand and regulations that require reclamation of land-filled FA [18]. However, its long-term growth potential is in question due to reductions in coal-combustion electricity production [2,16,19,20] and the high demand of FA coming from numerous applications, including soil amendment and structural fill [21]. The nonhomogeneous quality of FA that depends on a number of factors including the parent coal, plant operations, and post-combustion processes can further limit its use as a SCM.
Expected supply limitations of FA and GGBS in the near future are encouraging researchers to explore additional options; other minerals and waste materials can be used as SCM [22][23][24][25][26][27][28]; optimized low-clinker system can be used as a binder [29]; clinker-free geopolymer can replace OPC [30]; self-healing concrete can mitigate CO 2 emissions from repair events [31,32]; and ductile engineered composites can reduce life cycle CO 2 emission with an extended service life [33,34]. Another area of active research includes so-called 'CO 2 utilization' in concrete where concrete is formulated with added CO 2 either in its constituents before casting, during batching and mixing, or in finished products. Three main CO 2 utilization strategies in concrete formulation can be found in the literature as shown in figure 1, including carbonation curing [35][36][37], carbonation during mixing [38][39][40][41], and carbonation with recycled concrete aggregate (RCA) [42]. Thus far, the majority of studies on CO 2 utilization in concrete exclusively explored mitigation opportunities through sequestration using carbonation curing [8,[43][44][45][46][47]. By implementing this strategy worldwide, it is estimated that about 30-300 million tonnes (Mt) of CO 2 could be sequestered globally in the future [46], which is close to 1%-10% of CO 2 emitted from manufacturing cement [5].
A review of the literature indicates that most of the state-of-the-art research on CO 2 utilization has failed to consider the systems-level design optimization of concrete mixtures achievable with embodied CO 2 . When CO 2 is introduced to freshly cast concrete or into its constituents prior to casting, the introduced CO 2 either becomes part of the binding matrix or improves the mechanical properties of the constituents, resulting in increased overall concrete strength. Hence the same level of compressive strength can be achieved with a lower amount of binder, resulting in a significant decrease of CO 2 emissions. Indeed, the effectiveness of this strategy was shown in a recent case study when CO 2 was added during mixing, where the overall scale of CO 2 mitigation was magnified by 35 times compared to sequestration alone [39]. In this work, we conduct a more comprehensive assessment of this strategy. CO 2 utilization alone may not be an economical option for reducing CO 2 emissions in concrete compared to other alternatives, even when it is financially supported by tax credits such as 45Q that reward sequestered CO 2 . However, the saved material cost associated with reduced binder could help promote CO 2 utilization in a cost-competitive manner. For instance, the cost of abating one tonne of CO 2 Figure 1. Flow diagram of concrete mixture formulation. CO 2 can be utilized and permanently sequestered in RCA, during batching and mixing, or during the curing process. Chemical admixtures typically comprise less than 1% of the total mass of concrete and thus are excluded from this study (see section 3.1 in the supplementary material available online at stacks.iop.org/ERL/14/114014/ mmedia). emissions through diverse management and technological interventions, including on-site CCS in cement plants, has been estimated to cost between $5 and $450 [48]. However, offsetting one tonne of CO 2 by implementing carbonation curing could incur additional costs between $350 and $750 [49]. Given that the cost of manufacturing binders, especially OPC, constitutes over half of the total cost of concrete mixtures, the additional cost incurred by CO 2 utilization can potentially be mitigated if there is concomitant reduction in the use of binder materials. In this case, the concrete formulated with added CO 2 and reduced binder loading could be cost-competitive with conventional concrete.
In this study we assess the potential scale of net CO 2 reduction and net cost savings achievable by employing the combined strategy of reducing binder and utilizing CO 2 in concrete formulations. This study is conducted at the national level, utilizing a set of concrete mixtures that represents the diversity of formulations used in the US. Compressive strengths measured at 28 d are used as a primary technical criterion for all mixtures. We evaluate three CO 2 utilization strategies -including carbonation curing, carbonation during mixing, and addition of carbonated RCA-by employing each of them in combination with reduced binder at a time. We examine the overall CO 2 reduction potential of the combined strategy against direct CO 2 sequestration as a result of utilizing CO 2 alone.
The financial incentive of reducing binder loading is compared with the tax incentive associated with the 45Q tax credit as amended in the FUTURE Act bill.

Methods
Our national CO 2 footprint and material cost assessments of concrete manufacturing in the US are conducted at a mixture level, using a set of mixtures that collectively represents the industry. Figure 2 summarizes the derivation process of a CO 2 -amended mixture from a conventional mixture and how the collection of mixtures is used to emulate the US concrete industry. Here, concrete mixtures are plotted in terms of compressive strength and binder intensity, which is defined by normalizing total binder content in a mixture with its compressive strength as shown below. In this study, binders may include OPC, FA, GGBS, silica fume, metakaolin, and natural pozzolans that contain volcanic ash and shale ash. Limestone may also be used to substitute binders as a filler: ] binder intensity total binder kg m compressive strength MPa .   Since each D Alt has lost a fraction of its compressive strength relative to its original D Base , each D Alt is then formulated with added CO 2 to recover its lost compressive strength. The addition of CO 2 provides a boost in compressive strength to each D Alt , yielding a new set of mixture designs that we call D CO 2 , whose compressive strength is equal to or surpasses that of D Base , as shown in figure 2(A3). From the set of D CO 2 mixtures derived from each of the 48 D Base mixture designs, the one that features the lowest net CO 2 footprint-considering both the effects of reducing binder (avoiding CO 2 emissions in the formulation) and adding CO 2 (absorbing CO 2 content in the formulation) -is called D .
Min CO 2 Since we employ one CO 2 utilization method at a time, we generate three sets of D , Min CO 2 each containing 48 mixtures that replace 48 D Base mixtures when CO 2 is added during mixing, curing, or during the recycling of concrete aggregate. Each of these sets of 48 D Min CO 2 mixtures, each substituting its D Base counterpart without compromising strength, represents the diversity of possible alternatives and thus simulates the alternative US concrete industry optimized for reduced binder use enabled by the inclusion of CO 2 in its formulation. Further details of this step-wise approach are provided in section 2**** in the supplementary material. CO 2 footprint and material cost of each of the D Base and D Min CO 2 mixtures are then calculated by aggregating constituent-level values. This includes CO 2 emitted and cost incurred from manufacturing of each constituent, transporting it to concrete plants and preparing them as a final product-mixed concrete for ready-mixed application or finished precast products. We assumed industrial grade (purity99.5%) liquid byproduct CO 2 from ethanol, ammonia, or hydrogen plants would be used for the CO 2 utilization processes.
The additional CO 2 emitted and cost incurred during purifying, liquefying and transporting byproduct CO 2 to concrete plants as well as from applying CO 2 treatment to concrete were included in our calculation [46,[52][53][54]. The CO 2 emitted and material cost incurred nationally from manufacturing concrete in the US can then be estimate by summing CO 2 footprint and material cost of each of the 48 mixtures weighted by its national production volume. Thus, the net CO 2 mitigation and material cost savings achievable for each of the three CO 2 utilization strategies are determined by assessing the changes in total CO 2 footprint and material cost of the 48 D Min CO 2 mixtures compared to those of the original 48 D Base mixtures. The assumptions, data sources, and calculations applied are available in sections 3-4 in the supplementary material.
With this approach, we investigate three possible mitigation cases to address existing uncertainties on available concrete mixture designs. The assumptions and procedures outlined above define an alternative US concrete industry that utilizes CO 2 to create reduced binder formulations at full scale. Since this case does not consider future supply restrictions of FA and GGBS, the resulting CO 2 mitigation potential is likely an overestimation. Therefore, the calculations here represent a 'high mitigation case' that could be achieved with a novel SCM that does not compromise compressive strength of concrete and therefore can replace high proportions of binders without additional CO 2 emissions. We then create what we consider to be a more realistic mitigation case that accounts for future supply restriction of FA and GGBS and use this as our 'nominal mitigation case'. This case is generated by recreating three sets of D Min CO 2 mixtures using each of the CO 2 utilization strategies following the same procedure as above but when deriving D Alt , we only use mixtures that have equal or less FA or GGBS content than D Base for D Alt . Lastly, we present a 'low mitigation case' which represents a limiting scenario where only commercial mixture designs that are being used today are available in the future without further improvement. This design-restricted scenario can be emulated by reproducing three sets of D Min CO 2 mixtures as we only allow conventional D Base mixtures to serve as D Alt mixtures. The opportunity to reduce binder loading is then highly constrained. Consequently, we generate nine results by evaluating three CO 2 utilization strategies in three levels of mitigation cases. Many input variables and data used in this study were also found with a range of uncertainty in the literature. These variables included the annual concrete consumption rate in the US, the CO 2 emissions and costs associated with manufacturing each constituent, and the maximum market penetration rate attainable for each of the CO 2 utilization strategies. Three uncertainty levels were allowed for each of these four parameters, generating 81 uncertainty cases. These 81 cases can be applied to the nine results created above to generate 729 results. Since results generated with 81 uncertainty cases show consistent trends, we only report median results in the following section for clarity. The results generated with the full spectrum of uncertainty cases are presented in section 6 in the supplementary material.

Results and discussion
If reduced binder loading is jointly employed with CO 2 sequestration throughout the concrete industry, the potential scale of atmospheric CO 2 reduction increases by over an order of magnitude compared to partially offsetting the emissions solely through CO 2 sequestration in the concrete. Figure 3 shows that the amount of CO 2 emissions that can be offset solely through sequestration (horizontal dotted lines) is limited to about 1.09%-1.23% of current CO 2 emissions from the US concrete industry if CO 2 is added during mixing or in RCA. With respect to the possibility of reducing binder to its maximal potential in the nominal mitigation case, the industry-wide CO 2 mitigation potential increases to about 31%-38% of the emissions when CO 2 is added in RCA or during mixing. Considering that approximately 75.5 Mt of CO 2 are annually emitted from concrete manufacturing in the US, this translates into about 23.3-28.5 Mt of mitigated CO 2 where over 98% of which are avoided CO 2 emissions from reduced binder use (see section 7 in the supplementary material for further information). Compared to CO 2 offset through sequestration, this represents a 25×-35× increase in mitigated CO 2 emissions. Although more CO 2 is sequestered in carbonation curing per volume of treated concrete, the total potential to reduce industry-wide CO 2 emissions is relatively low (about 4% of total industry emissions), as this approach is limited to the precast industry. The potential scale of CO 2 reduction when carbonation curing is employed would likely be smaller considering that carbonation curing can only treat the volume of concrete near the surface and the actual precast products tend to have a larger volume than previously studied lab samples.
The differences between low, nominal, and high mitigation cases mainly result from assumptions around the capacity to reduce OPC content from mixtures in each case, which is primarily determined by the abundance and diversity of the SCM available to replace OPC. For instance, the high mitigation case assumes it is possible to increase the use of GGBS by about six to eight times compared to the current level, which reduces OPC consumption by over 60% when CO 2 is added during mixing or in RCA. Consequently, the mitigation potential can expand up to 54 times compared to the CO 2 reduction achievable with direct sequestration alone. However, when the incorporation rate of GGBS is constrained below the current level in the nominal case, OPC is reduced primarily by expanding the use of limestone, which is a trend consistent with observations in previous literature [2,55]. In the nominal cases shown in figure 3, about 11%-19% of limestone by mass of OPC is included in mixtures when CO 2 is added during curing or mixingwhich is comparable or higher than the typical incorporation rate of up to 15% and made possible by CO 2 -induced compressive strength enhancement [28,56]. As a result, about 32%-41% of OPC is replaced. It is the adverse impact on compressive strength from dilution that prevents further replacement of OPC with limestone, limiting additional binder reduction. When the available mixture design for D Alt is further restricted to existing benchmark designs in the low mitigation case, limestone is no longer available and mitigation potential is further constrained. But even in this restricted case, CO 2 mitigation potential may increase by nearly an order of magnitude compared to partially offsetting emissions through CO 2 sequestration alone if reducing binder is considered.
Theoretically, the three CO 2 utilization technologies investigated in this study can be jointly implemented to maximize CO 2 mitigation as they target different components of concrete. For instance, He et al found a compound boost in compressive strength when CO 2 was added during mixing and also during curing [57]. The compound effect may prove important in the future when optimizing concrete formulations. However, few experimental studies to date have investigated nonlinear interactions between carbonated components and their impact on compressive strength. Therefore, joint application of multiple CO 2 utilization techniques was not considered here and the presented results are based on implementing one CO 2 utilization technique at a time.
In the nominal mitigation cases, maximal OPC reductions range from 32%-41%. Since OPC comprises nearly 50% of the material cost, this results in a total cost reduction of 18% or $3.6 billion across the US if CO 2 were added universally during mixing or in RCA ( figure 4). Cost reduction is limited to 1.5% or $300 million when carbonation curing is implemented nationwide. In figure 4, these cost savings are superimposed with the cost uncertainty of CO 2 utilization, which includes both the cost of byproduct CO 2 and the cost of carbonation treatments. The byproduct CO 2 is assumed to cost between $34/tCO 2 and $69/tCO 2 and the cost of CO 2 utilization processes ranges between $10/tCO 2 and $2000/tCO 2 (see sections 3.2-3.3 in the supplementary material for details). The upper end of the cost uncertainty reflects non-optimized cases and thus the realistic cost is expected to be closer to the lower end of the spectrum. The cost incurred by utilizing CO 2 nationally is then estimated by multiplying these CO 2 -normalized cost bounds with the total amount of CO 2 sequestered in concrete throughout the industry. The yellow bounds in figure 4 depict the range of cost of utilizing CO 2 nationally.
We find that the saved material cost could fully mitigate the additional cost of CO 2 utilization without external financial support if OPC content in concrete is reduced by either over 5% when CO 2 is added in RCA or beyond 16% when CO 2 is injected during mixing. In these cases, the resulting D Min CO 2 mixtures would have the same cost as their D Base counterparts. Figure 4 shows that if OPC use can be reduced further, the saved cost (red curve) clearly rises above the cost uncertainty of CO 2 utilization (yellow range) and there is a strong possibility that the D Min CO 2 mixtures cost less than their D Base counterparts. For instance, when CO 2 is added during mixing or in RCA with maximal OPC reduction, over $7000 is saved per tonne of CO 2 sequestered. This is more than enough to mitigate the additional cost incurred by utilizing CO 2 , which can range between $34 and $2070 per tonne of CO 2 utilized ( figure 4). If the saved cost is contained within the cost uncertainty-as in the case of carbonation curing, or when CO 2 is added during mixing or in RCA with a low degree of OPC reduction-additional support might be needed to make D Min CO 2 mixtures cost-competitive. As an example, D Min CO 2 mixtures prepared with carbonation curing would require up to $210/tCO 2 of carbon credit to cost the same as its D Base counterparts when its OPC reduction is maximized to about 38% ( figure 4(C)).
In the US, the amended 45Q provides up to $35 of tax credit for every tonne of CO 2 sequestered as a result of beneficial use. The applicable tax credit for concrete products will be lower if CO 2 is sourced from external entities as the credit needs to be split between concrete manufacturers and CO 2 capture facilities, the recipient of the credit under 45Q. However, the minimum sequestration requirement of 25000 tCO 2 per year may bar many concrete manufacturers from participating in 45Q. The concrete industry is comprised of highly distributed small-scale operators to accommodate local demands; the average market share of a typical ready-mixed and precast facility is approximately 0.018% and 0.093%, respectively [58,59]. As shown in figure 5(A), the amount of CO 2 that can be sequestered in concrete by a single concrete facility significantly falls short of the minimum sequestration requirement even when all of its concrete is fabricated with added CO 2 . Instead, all concrete mixtures produced by a leading manufacturer may need to be formulated with added CO 2 across all its facilities in the US to be eligible for 45Q. An example of a leading ready-mixed company shown in figure 5(B) operates 335 locations in the US with a market share of 4.7% in ready-mixed industry [58]. An example precast leader has 77 locations in the US with a market share of 9.1% in precast industry [59]. Such a high initial eligibility criterion may defeat the purpose of 45Q to foster the implementation of emerging CO 2 technologies. Additionally, the sequestration-based mechanism of 45Q does not incentivize binder reduction since the reduced CO 2 emissions from avoided binder use are not applicable for the credit. As a result, 45Q may distract industry attention away from CO 2 utilization methods that do not sequester a high proportion of CO 2 , such as CO 2 addition during mixing or in RCA, even though such techniques have a much higher net carbon mitigation potential. This disproportional incentive is also shown in figure 5; it is relatively easier to meet the eligibility criterion of 45Q with carbonation curing than with other CO 2 utilization methods. Hence, a more comprehensive incentive system should encourage binder reduction strategies to be coupled with CO 2 addition. Such a mechanism could promote material cost savings while driving aggressive CO 2 emission reductions.
Although this analysis is based on a simplified technical criterion of 28-day compressive strength and with a primary focus in the US, the main findings that more efficient use of OPC provides significant CO 2 reduction opportunity and that added CO 2 can catalyze further OPC reduction would be valid for the global concrete industry. To illustrate this, the results generated with randomly selected D Base mixtures from the database indeed show a consistent trend with the results created with the original 48 D Base mixtures (see section 8 in the supplementary material). But other parameters excluded in this analysis such as workability, set time, permeability, early compressive strength and compressive strength measured at a longer period beyond 28 days need to be considered when selecting alternative mixtures for practical purposes. The specific requirement for fresh or hardened concrete properties, and potential synergies and tradeoffs between added CO 2 and constituents could influence the final mixture choices [37,39].
Future studies on CO 2 utilization in concrete need to assess the impact of added CO 2 on ductility and tensile strength, which dictate long-term durability and the life cycle CO 2 footprint of concrete rather than compressive strength; past efforts to strengthen concrete structures by increasing their compressive strength has led to premature deterioration in recent decades [60]. This is because increasing compressive strength tends to make concrete more vulnerable to cracking [60], and the ingress of external detrimental species along the cracks ultimately deteriorates structural integrity. The resulting additional repair events and reduced service life can substantially increase the life cycle CO 2 emissions and cost of concrete. For instance, comparative studies on a bridge deck showed nearly 30% reduction in life cycle CO 2 emissions and cost when brittle concrete was replaced with engineered cementitious composite (ECC) that has enhanced ductility by several orders of magnitude [33,61]. If the added CO 2 in ECC can be used to reduce OPC content without compromising its ductility, additional CO 2 emission may be mitigated. But the review of existing literature shows lack of experimental data beyond compressive strength; among the 696 collected mixtures used in this study, only about 9% were measured for tensile strength and none were tested for ductility; none of the carbonated mixtures were evaluated for their impacts on tensile strength or ductility. This lack of data highlights both challenges and future opportunities for CO 2 utilization.

Conclusion
Despite the recent interest in devising an actionable and climate-compatible CO 2 reduction roadmap in concrete industry, the combined strategy of reduced binder use and CO 2 utilization has not been recognized enough a highly desirable option [2,6,62]. This is a critical concern as the global implementation of the combined strategy would enable gigatonne-scale CO 2 emissions reduction without relying on novel SCMs, alternative binders, or CCS. For instance, if the nominal mitigation potentials determined in this study were scaled to a global level, roughly 930-1100 Mt of annual CO 2 emissions from the concrete industry could be mitigated by incorporating CO 2 in concrete formulation during mixing or in RCA. If carbonation curing were applied worldwide, about 110 Mt of CO 2 emissions could be mitigated. Comparatively speaking, only about 32-41 Mt of CO 2 could be mitigated globally solely through direct sequestration. Therefore, we find that the combined strategy of binder reduction and CO 2 utilization could provide economically viable, aggressive CO 2 reduction pathways that can help advance the concrete industry toward carbon neutrality.