Synergistic Utilisation of Construction Demolition Waste (CD&W) and Agricultural Residues as Sustainable Cement Alternatives: A Critical Analysis of Unexplored Potential

Abstract

Decarbonising the construction industry’s substantial ecological footprint demands credible substitutes that preserve structural performance while valorising waste. Although construction and demolition waste (CD&W) has been widely studied, the vast potential of agricultural residues (e.g., corncob, rice husk) and, crucially, their synergy remains underexplored. This study couples a systematic literature review with mathematical modelling to evaluate binary CD&W–agro-waste binders. A modified Andreasen–Andersen packing framework and pozzolanic activity indices inform multi-objective optimisation and Pareto analysis. The optimum identified is a 70:30 CD&W-to-agricultural ratio at 20% total cement replacement, predicted to retain 86.0% of OPC compressive strength versus a 79.4% average for single-waste systems (8.3% non-additive uplift). Life-cycle assessment (cradle-to-gate) shows a 20.3% carbon reduction for the synergistic blend (vs. 19.6% CD&W-only; 19.3% agro-only); when normalised by strength (kg CO₂-eq/MPa·m³), the blend delivers 6.3% better carbon efficiency than OPC (5.63 vs. 6.01), outperforming agro-only (5.79) and CD&W-only (6.61). Global diversion arithmetic indicates feasible redirection of 0.246 Gt y⁻¹ of wastes (5.7% of CD&W and 1.8% of agricultural residues) at 30% market penetration. Mechanistically, synergy arises from particle size complementarity, complementary Ca–Si reactivity generating additional C–S–H, and improved rheology at equivalent flow. Monte Carlo analysis yields a 91.2% probability of ≥40 MPa and 78.3% probability of ≥80% strength retention for the optimum; the 95% interval is 39.5–55.3 MPa. Variance-based sensitivity attributes 38.9% of output variance to the Bolomey constant and 44% to pozzolanic indices; interactions contribute 19.5%, justifying global (not local) uncertainty propagation. While promising, claims are bounded by cradle-to-gate scope and the absence of empirical durability and end-of-life evidence. The results nevertheless outline a tractable pathway to circular, lower-carbon concretes using co-processed waste. The approach directly supports circular economy goals and scalable regional deployment.

1. Introduction

The accelerating pace of global urbanisation and infrastructure development has intensified scrutiny of the construction industry’s environmental footprint, which accounts for approximately 11% of global CO₂ emissions [1]. Within this context, cement production is particularly problematic, contributing 8% of worldwide carbon emissions through energy-intensive manufacturing processes that combine high-temperature calcination with substantial resource consumption [2,3]. As the second most consumed material on earth after water, cement’s environmental impact extends beyond carbon emissions to encompass significant resource depletion and waste generation challenges that demand urgent attention in an era of heightened climate awareness [4,5]. The magnitude of waste generation across the construction and agricultural sectors presents both environmental challenges and unprecedented opportunities for sustainable material development. While the construction sector generates approximately 35% of global solid waste through demolition and renovation activities [6], agricultural operations simultaneously produce billions of tons of organic residues annually which creates parallel waste management crises [7]. This situation, where two of the world’s largest industries generate substantial waste streams, presents a compelling opportunity to transition from linear “take–make–dispose” models toward circular systems that transform waste into valuable resources. Rather than viewing these waste streams as environmental burdens, innovative cross-sectoral collaborations can unlock their potential as feedstock for new products and processes, creating closed-loop systems that eliminate waste while generating economic value. This shift from linear disposal to circular resource flows not only addresses environmental challenges but also opens pathways for regenerative business models that continuously recycle materials back into productive use.

Contemporary research in sustainable construction materials has predominantly pursued silo approaches, investigating either CD&W [8] or agricultural residues [9] as individual cement alternatives without considering their synergistic potential. This recognises the fact that many regions experiencing rapid construction demolition and waste also maintain substantial agricultural production, presenting frequent opportunities for integrated waste management and innovative material development strategies. The geographic coincidence of high construction demolition and wastes with abundant agricultural production suggests that coordinated utilisation of both waste streams could achieve superior environmental and economic outcomes compared to previously adopted solo approaches to waste management. The theoretical foundation for synergistic material combinations in construction is well-established through historical precedents, most notably the successful integration of fly ash and slag in concrete production [10]. These binary and ternary blended systems have consistently modelled potential performance characteristics compared to individual supplementary cementitious materials (SCMs), achieving enhanced strength development, improved durability, and reduced environmental impact through complementary chemical and physical interactions [11]. This proven track record of synergistic benefits in cementitious systems provides compelling evidence that similar or enhanced performance could be achieved through a strategic combination of CD&W and agricultural residues. The technical rationale for CD&W–agricultural residue synergies lies in their complementary material characteristics and the potential to address individual limitations through combined utilisation. Construction demolition waste typically contributes aggregate-like properties and moderate pozzolanic activity, while agricultural residues (particularly corn cobs and rice husks) offer high silica content and ultrafine particle sizes that enhance pozzolanic reactivity. These contrasting yet complementary properties suggest that strategic blending could optimise particle size distribution, enhance chemical reactivity, and improve overall system performance while simultaneously addressing waste management challenges across multiple sectors.

Despite this compelling theoretical foundation, the literature reveals critical gaps in systematic investigation of CD&W–agricultural waste synergies. Current research efforts remain fragmented, focusing on single waste materials and limiting our understanding of potential interactive effects, optimised blending strategies, and performance enhancement mechanisms. This oversight is particularly problematic given the urgent need for scalable sustainable construction materials and the growing recognition that circular economy principles require integrated approaches that maximise resource utilisation efficiency across multiple waste streams. The absence of comprehensive frameworks for evaluating multi-waste material systems further compounds these research limitations, creating barriers to systematic comparison of different synergistic combinations and optimisation of their development for specific applications. Standard testing protocols, performance criteria, and sustainability assessment methods designed for individual SCMs prove inadequate for complex multi-waste systems that exhibit novel interactive behaviours and performance characteristics.

Addressing these research gaps requires a fundamental shift from isolated waste stream investigation to integrated synergistic approaches that recognise the interconnected nature of waste generation and resource utilisation challenges. This study responds to this need by critically examining the unexplored potential of CD&W–agricultural residue combinations, with particular focus on corn cobs and rice husks: two of the most abundant agricultural residues globally with minimal nutritional value and significant disposal challenges [12]. Through comprehensive literature analysis and mathematical modelling, this research investigates whether synergistic combinations of these waste streams can create more effective and sustainable cement alternatives compared to individual waste utilisation, while simultaneously contributing to policy development for integrated waste management strategies that address environmental challenges across the construction and agricultural sectors. However, it should be noted that the current study is limited by the absence of direct experimental work. Nonetheless, this is balanced by the comprehensive synthesis of existing knowledge and the development of theoretical frameworks that can guide future empirical research. This approach is particularly valuable in identifying the most promising research directions and avoiding duplication of existing efforts. It should be explicitly noted that the current study relies solely on theoretical modelling and mathematical optimisation. Therefore, the predicted synergistic benefits require thorough experimental validation before practical application.

This study advances the field along three fronts: First, it integrates construction and demolition fines with silica rich agricultural ashes and shows how their different roles can be balanced in one binder. It presents a theoretically optimal 70:30 CD&W/agricultural waste ratio at 20% total replacement rather than treating waste in isolation as most prior reviews do. Second, the analysis links packing to the progress of hydration through the starting porosity and uses a calibrated strength mapping, then reports probabilities of meeting target strengths rather than single point values—thereby guiding experimental calibration. Third, the environmental assessment compares carbon per unit of strength and volume, considers regional electricity mixes and future grid decarbonisation, and finds that the combined waste system keeps a small but consistent advantage over ordinary cement. Together, these factors establish a distinct research contribution of integrated waste-to-binder design, mechanistically justified hybrid modelling, probabilistic performance assessment, and decision-oriented environmental metrics.

2. Literature Review

2.1. Environmental Impact of Cement Production

Cement production is inherently carbon-intensive, requiring high-temperature calcination processes that release CO₂ both from fuel combustion and limestone decomposition [13]. The industry faces mounting pressure to reduce its environmental footprint, with the Paris Agreement requiring 16% emissions reduction by 2030 relative to 2015 levels [14]. Traditional approaches to emissions reduction, including energy efficiency improvements and alternative fuels, have proven insufficient to meet these targets, necessitating fundamental changes in cement composition and production methods. The embodied energy of cement ranges from 4.2 to 7.5 GJ/tonne, significantly higher compared to most other construction materials [15]. This high energy intensity, combined with the vast quantities required for global construction activities, underscores the critical need for sustainable alternatives that can maintain structural performance while reducing environmental impact.

2.2. Construction Demolition Waste as Cement Alternative

Construction demolition waste represents a substantial and growing waste stream, with global generation estimated at over 3 billion tonnes annually [16]. CD&W typically contains concrete, masonry, metals, wood, and gypsum, with concrete and masonry comprising 70–80% of the total volume [17]. The recycling potential of these materials has been recognised, with recycled concrete aggregates becoming increasingly common in new construction applications. Research on CD&W-based cement alternatives has shown promising results. Crushed concrete waste has demonstrated pozzolanic properties when processed appropriately, contributing to strength development in cementitious systems [18]. However, challenges include variability in composition, potential contamination, and the need for extensive processing to achieve suitable particle size distributions and remove deleterious materials [19]. Recent studies have explored various CD&W fractions as partial cement replacements. Masonry waste, when finely ground, exhibits pozzolanic activity due to its clay brick content which contributes to long-term strength development [20]. Similarly, recycled concrete fines have shown potential as supplementary cementitious materials, though their effectiveness varies significantly based on the original concrete quality and processing methods [21]. Recent reviews quantify these effects and outline activation strategies [22] improving the reactivity of industrial recycled concrete fines (hydrothermal/mechanical activation; carbonation), with strength and microstructure gains [23]. A cross-material comparison of reactivity, fineness requirements, optimal dosages, and durability outcomes is summarised in

Table 1. Comparative synthesis across CD&W fines and agricultural ashes.

Material (Binder Replacement)	Dominant Reactivity and Ca/Si Effect	Fineness/Processing Notes	Optimal Replacement (Mass of Cement)	Durability Outcomes Near Optimum
Recycled concrete fines (RCF, untreated)	Mainly filler/nucleation; limited intrinsic pozzolanicity; minor Ca/Si change	Grinding helps dispersion; activation often required for chemical reactivity	≤~10%	Mixed to neutral; improvements generally not reported without activation; transport properties variable
Carbonated recycled concrete powders/fines (CRCF/RCPP)	Enhanced reactivity after carbonation; denser microstructure; modest Ca/Si reduction via secondary products	Require complete carbonation and fine grinding; process control is critical	≤~20% advised in recent mortar/paste studies (activity ↑, strength/durability acceptable)	Strength and chloride transport often improve versus untreated RCF; benefits depend on carbonation method and mix design
Rice husk ash (RHA)	High amorphous SiO2 → strong pozzolanicity; lowers C–S–H Ca/Si; pore refinement	Controlled combustion + fine grinding essential	~10–40% widely supported	Chloride ingress ↓; freeze–thaw performance ↑; carbonation trend depends on curing/strength equivalence
Wheat straw ash (WSA)	Silica-rich; reactivity sensitive to calcination; Ca/Si trend similar to RHA when well processed	Calcine ~600–700 °C; grind to raise amorphous fraction	~5–10%	Mechanical and several durability metrics improve at low dosages; outcomes sensitive to alkalis/processing
Biomass fly ash (BFA)	Moderate pozzolanicity (source-dependent); Ca/Si of C–S–H varies with chemistry	Classification to control LOI/alkalis; adequate fineness	~10–20% typical	Reports of reduced chloride migration and acceptable durability after processing; variability is high
Corncob ash (CCA)	Silica-bearing pozzolan; benefits at modest dosages; contributes to lower Ca/Si C–S–H when calcined/ground	Pretreatment (e.g., washing to reduce K2O), controlled calcination, fine grinding improve reactivity	~10–30% recommended; several reviews advise ≤10%	At optimum: strength retention and RCPT reductions reported vs. plain mixes; higher dosages risk strength loss if unprocessed

.

2.3. Agricultural Residues in Cement Production

Agricultural residues, particularly those rich in silica, have attracted attention as potential cement alternatives due to their pozzolanic properties. Rice husk ash (RHA), produced by controlled combustion of rice husks, contains 85–95% silica in amorphous form, making it highly reactive in alkaline environments [24]. Studies have demonstrated that RHA can replace up to 20% of cement without compromising concrete strength, while improving durability properties [25]. Indumathi, M. et al. [26]’s synthesis work reaffirms these ranges and durability trends. Corncob ash has similarly shown promise as a pozzolanic material. Research by Okeke et al. [27] demonstrated that corncob ash could replace up to 15% of cement in concrete applications, with proper grinding and calcination. The silica content of corncob ash typically ranges from 60 to 70%, lower than RHA but still sufficient for pozzolanic reactions [27]. Olive waste ash (OWA) has emerged as a particularly promising agricultural residue, driven by substantial waste generation in Mediterranean olive oil production, with recent comprehensive studies by Ghazzawi et al. [28] demonstrating optimal replacement levels of 10% while exhibiting both cementitious and pozzolanic behaviour. Sugarcane bagasse ash (SCBA) represents one of the most extensively studied agricultural residues, with França et al. [29] conducting comprehensive feasibility studies at replacement rates of 10–30%, demonstrating that processed SCBA achieves optimal performance when ground and re-burnt, while recent investigations by Althaqafi et al. [30] have shown that combining SCBA with metakaolin and polypropylene fibres can enhance compressive strength by 7–9% at 15% metakaolin incorporation levels. Wheat straw ash (WSA) has gained attention as an alternative to environmentally harmful open-field burning practices, with controlled calcination at 600 °C producing superior pozzolanic activity [31], Recent testing confirms WSA’s viability and processing–performance links [32], while response-surface methodology optimisation by Bheel et al. [33] identified 10.12% WSA content as optimal for achieving maximum performance. Coconut shell ash (CSA) applications have expanded to tropical regions, with recent research establishing 15–20% replacement levels for maintaining structural performance while achieving 15% reduction in global warming potential [34]. Other agricultural residues investigated include groundnut ash, palm kernel ash, and corn stover ash. While results vary based on processing conditions and ash composition, most studies report successful partial cement replacement levels of 8–25% [35]. However, challenges include seasonal availability, transportation costs, and the need for controlled combustion to achieve optimal pozzolanic properties.

2.4. Synergistic Material Combinations in Construction

The concept of synergistic material combinations is well-established in construction materials science. Binary and ternary blended cements, incorporating materials such as fly ash, slag, and silica fume, have demonstrated predicted improvement compared to individual SCMs [36]. These synergistic effects typically result from complementary particle size distributions, varied chemical compositions, and different reaction kinetics. Successful synergistic combinations often address individual material limitations. For example, the combination of fly ash and slag in concrete provides improved workability from fly ash and enhanced early-age strength from slag [37]. Recent work on multi-waste binders further evidences synergy through complementary chemistries and PSDs [38]. This precedent suggests that combining CD&W and agricultural residues could similarly address complementary limitations while maximising individual strengths.

2.5. Comparative Mechanisms and Performance Synthesis for CD&W Fines and Agricultural Ashes

From the preceding subsections, a mechanistic commonality (filler/nucleation effects and secondary C–S–H formation from amorphous silica) and key differences (intrinsic reactivity, Ca/Si adjustment in C–S–H, fineness sensitivity, practical replacement windows, and durability trade-offs) across CD&W fines and agricultural ashes are illustrated.

• Common mechanisms across systems

  (a)

  Filler and nucleation (physical): Finely divided particles accelerate early hydration and refine pore structure irrespective of intrinsic pozzolanicity—especially relevant to recycled fines and lower-reactivity ashes.

  (b)

  Secondary C–S–H formation (chemical): silica-rich ashes (RHA/WSA/BFA/CCA) consume portlandite and form additional C–S–H that typically exhibits a lower Ca/Si than clinker-derived C–S–H, densifying the microstructure and improving transport performance [39].

• Key differences guiding mixture design

  (a)

  Untreated recycled concrete fines (RCFs) mainly act as fillers with limited intrinsic pozzolanicity; carbonation-activated recycled powders/fines (CRCFs/RCPPs) show higher reactivity and enable practical replacement levels when properly processed. Recent work on carbonated recycled powder suggests not exceeding ~20% replacement in mortar/paste to avoid strength loss while still gaining activity and durability benefits [40].

  (b)

  Rice husk ash (RHA): For high amorphous SiO₂, there is robust evidence for ~10–20% optimal replacement in concrete with durability benefits (chloride and freeze–thaw), contingent on controlled burning and fineness [41].

  (c)

  Wheat straw ash (WSA): Reactivity depends strongly on calcination regime and fineness; recent studies show viable performance at ≈5–10% with durability gains when properly processed [42].

  (d)

  Biomass fly ash (BFA): Composition is source-dependent; with classification/finishing, 10–20% replacement is common, with several studies reporting reduced chloride migration and acceptable durability [43].

  (e)

  Corncob ash (CCA): For silica-bearing agricultural ash, recent reviews and experiments indicate the best performance around 5–10%, with several sources recommending ≤10% to avoid penalties; pretreatment (washing to reduce K₂O), adequate calcination, and fine grinding improve strength and durability (e.g., lower RCPT) [44].

The lower Ca/Si of C–S–H produced from silica-rich SCMs (including RHA/WSA/CCA/BFA) underpins reduced permeability/ion transport and aligns with observed durability trends.

These comparisons clarify why silica-rich agricultural ashes (RHA/WSA) are expected to contribute more to chemical densification (lower Ca/Si and tighter pore network), whereas CD&W fines contribute chiefly via packing/nucleation unless activated (e.g., carbonation, hydrothermal/mechanochemical treatment). Consequently, the most promising blends are those that combine a reactive silica source (RHA/WSA/BFA) with properly processed CD&W fines that supply nucleation sites and packing benefits—aligning with the current study’s synergy premise.

3. Methodology

3.1. Research Design

This study employs a mixed-methods research approach combining systematic literature analysis, comparative material performance, and environmental impact assessment. The methodology is designed to provide a comprehensive understanding of CD&W–agricultural waste synergies as it focuses on material performance and large-scale environmental impact.

3.2. Literature Analysis Framework

A systematic literature review was conducted to identify and analyse existing research on CD&W and agricultural residues as cement alternatives. The search strategy includes peer-reviewed journal articles, conference proceedings, and technical reports published in databases including Web of Science, Scopus, and Engineering Village. Search queries used specific keyword combinations related to cement alternatives, construction waste, and agricultural residues. Inclusion criteria for literature selection include the following: (1) studies focusing on CD&W or agricultural residues as cement/concrete alternatives, (2) peer-reviewed publications in English, (3) studies providing quantitative performance data, (4) research published within the specified time frame. Exclusion criteria include the following: (1) studies focusing solely on aggregates without cementitious properties, (2) non-peer-reviewed sources, (3) studies without sufficient technical detail for analysis.

3.3. Material Assessment Protocol

The material performance component focused on analysing published data and developing theoretical frameworks for understanding CD&W–agricultural waste interactions, while providing valuable insights into material behaviour and combined performance potential. Theoretical blend designs using mathematical models were developed to predict optimal blending ratios based on published individual material properties. The developed mathematical framework provides a systematic approach for predicting and optimising synergistic waste combinations through integrated particle packing and pozzolanic activity modelling. The modified Bolomey equation successfully incorporated multiple performance factors, including packing density (PF), pozzolanic activity index (PAI), and synergy factor (SF), to predict compressive strength development. The framework’s strength lies in its systematic consideration of material interactions through established theoretical relationships, building upon foundational work by Powers and Brownyard [45] for cement hydration modelling.

3.3.1. Theoretical Framework and Model Integration Justification

The integration of the modified Bolomey equation with particle packing models and pozzolanic activity indices requires careful theoretical justification to avoid the pitfall of inappropriately merging empirical and mechanistic frameworks. This study addresses this concern through a hierarchical modelling approach that maintains physical consistency while acknowledging the inherent limitations of theoretical predictions without experimental calibration.

• Bolomey Equation: Empirical Foundation and Mechanistic Interpretation

The Bolomey equation, originally proposed by Bolomey [46], represents one of the earliest attempts to relate concrete strength to the cement-to-water ratio through the following empirical relationship:

f_c = K(C/W − 0.5)^n

While empirical in origin, this equation has a sound mechanistic basis rooted in Powers and Brownyard’s gel–space ratio theory [47], which established that concrete strength is fundamentally governed by the ratio of hydration products (gel) to available space (gel plus capillary pores). The constant “K” and exponent “n” in the Bolomey equation are calibration parameters that implicitly capture the intrinsic strength of the cement gel and the relationship between porosity and strength, respectively [47].

• Modification for Supplementary Cementitious Materials (SCMs)

The extension of the Bolomey equation to systems containing supplementary cementitious materials requires modification of the effective cement-to-water ratio to account for pozzolanic contributions. This approach is well-established in the concrete technology literature, with Papadakis [48] demonstrating that the pozzolanic activity index (PAI) provides a rational basis for calculating equivalent cement content:

(C/W)_eff = (C + Σ(w_i × PAI_i))/W

This modification is not arbitrary but reflects the fundamental mechanism by which pozzolanic materials contribute to strength development through secondary hydration reactions with calcium hydroxide to form additional calcium–silicate–hydrate (C-S-H) gel [49]. The PAI thus serves as a reactivity coefficient that weighs the contribution of each SCM relative to ordinary Portland cement.

• Integration of Particle Packing Effects

The particle packing factor (PF) in the modified equation represents the influence of particle size distribution on the initial capillary porosity of the paste, which directly affects the gel–space ratio at any given degree of hydration. This integration is theoretically justified through the sequential nature of concrete property development: (1) mixing establishes the initial particle arrangement and void structure (governed by packing), (2) hydration progressively fills these voids with reaction products [50].

The packing density (φ) calculated from the Andreasen–Andersen model with interaction coefficients [51] determines the minimum paste volume required to fill inter-particle voids, thereby establishing the starting capillary porosity. This initial porosity state then serves as the boundary condition for the hydration process modelled through Powers–Brownyard relationships. The particle packing factor is introduced as

PF = (φ_blend/φ_reference)^α

where φ_blend is the packing density of the blended system, φ_reference is the packing density of the control (OPC-only) system, and α is an exponent (typically 0.5–1.0) representing the sensitivity of strength to packing improvements [52]. This study adopts α = 1.0 as a conservative first approximation, acknowledging that the actual relationship may be non-linear and requires experimental validation. This linear assumption means that a 10% improvement in packing density translates to a 10% improvement in the packing factor contribution, which may overestimate or underestimate the true effect depending on the specific material interactions.

• Synergy Factor

The synergy factor (SF = 1.03) represents the performance enhancement beyond simple additive effects when multiple SCMs are combined. This factor accounts for interaction mechanisms that are not captured by independent consideration of packing and pozzolanic effects, including the following:

• (a)

  Chemical synergy: Complementary pozzolanic reactions where one SCM’s hydration products enhance another’s reactivity. For example, calcium-rich phases from recycled concrete powder can react with highly reactive silica from agricultural ashes to form additional C-S-H gel beyond what each material would produce independently [53].

  (b)

  Microstructural refinement: Complementary particle sizes create denser packing at multiple length scales, reducing the tortuosity of capillary pores and enhancing the effectiveness of pozzolanic reactions through improved spatial distribution of reactive sites [54].

  (c)

  Workability-mediated effects: Improved particle size distribution enhances paste rheology, potentially allowing reduced water content while maintaining workability, thereby improving the effective water-to-binder ratio [55].

The specific value of SF = 1.03 (representing a 3% enhancement) is derived from meta-analysis of published studies on binary and ternary blended cement systems. Matos et al. [56] reported synergistic strength improvements of 2–5% for binary slag–silica fume systems compared to rule-of-mixture predictions. Similarly, Qian and Li [57] documented 3–7% enhancements in ternary fly ash–slag–cement systems, attributing the effect to complementary reaction kinetics.

In this form, the framework is intentionally used as a pre-experimental decision-support tool rather than as a fully calibrated design model. Three choices keep the underlying assumptions transparent and anchored to the available evidence. First, all input ranges for pozzolanic activity indices, particle sizes, and strength–porosity parameters are bounded by the experimental variability synthesised in the literature review (Section 4.1), so that simulations do not extrapolate beyond conditions already observed for CD&W and agricultural residues. Second, the synergy factor is restricted to a narrow interval around unity, consistent with the modest (5–12%) strength uplifts reported for binary supplementary cementitious systems, and is treated explicitly as epistemic uncertainty rather than as a fixed “tuning” parameter. Third, mechanical predictions are interpreted in probabilistic terms against typical structural strength-retention thresholds (Section 4.2.4), rather than as point estimates. The subsequent results and discussion (Section 4.2.3, Section 4.2.4, Section 4.2.5 and Section 6) therefore regard the model outputs as quantitatively bounded hypotheses that guide, but do not replace, the targeted experimental programme.

3.3.2. Physically Consistent Coupling of Particle Packing and Hydration Models

The Andreasen–Andersen (A–A) particle packing model and the Powers–Brownyard (P–B) hydration framework describe different but complementary stages of paste formation. Packing determines the initial spatial arrangement of solids and interstitial voids at mixing; hydration then consumes water and generates solid hydration products that progressively fill the space. To maintain internal consistency, the current study coupled the models through the initial capillary porosity (or excess water) which is the natural state variable linking particle geometry to hydration kinetics.

• Step 1—Packing → initial void volume and water demand.

From the A–A distribution (with de Larrard’s interaction corrections), we compute the maximum paste-scale solid fraction ϕₚ (sometimes called “virtual packing density”). For a given solid volume $V_s$, the interstitial void volume that must be filled by fluid at t = 0 is

$$V_{\mathrm{void},\min }=V_s({\displaystyle \frac{1}{{\varphi }_p}}-1).$$

Given the batching water volume $V_{w,0}$ (i.e., specified workability), the initial capillary porosity is

$$V_{\mathrm{cap},0}=\mathrm{m}\mathrm{a}\mathrm{x} [ 0, V_{w,0}-V_{\mathrm{void},\min } ].$$

Physically, ϕₚ (hence $V_{\mathrm{void},\min }$) depends on the full particle size distribution of the binder (cement + CD&W fines + agricultural ash) and their interactions; it is not a strength multiplier but a space-filling constraint that sets the starting porosity and effective water film around particles. This interpretation aligns with the mixture-design literature where packing controls minimum paste volume, water film thickness, and water demand [51].

• Step 2—Hydration consumes capillary water and forms gel (P–B).

Hydration is then tracked with the P–B framework via the degree of hydration, α(t) [58]. For a cement mass $C$, the volumes of (i) chemically bound water, (ii) gel solids, and (iii) residual capillary water evolve as

$$V_{\mathrm{gel}}\left(t\right)={\nu }_{\mathrm{gel}} C \alpha \left(t\right), \Delta V_{w,\mathrm{cons}}\left(t\right)={\nu }_w C \alpha \left(t\right),$$

$$V_{\mathrm{cap}}\left(t\right)=V_{\mathrm{cap},0}-\Delta V_{w,\mathrm{cons}}\left(t\right) \left(\mathrm{floored} \mathrm{at} 0\right),$$

where ${\nu }_{\mathrm{gel}}$ and ${\nu }_w$ are P–B stoichiometric volume coefficients. Strength is linked to the gel–space ratio, $g(t)={\displaystyle \frac{V_{\mathrm{gel}}\left(t\right)}{V_{\mathrm{gel}}\left(t\right)+V_{\mathrm{cap}}\left(t\right)}}$, through the classical form $f_c\left(t\right)=A g(t{)}^m$ (A and m calibrated to the control). By construction, packing influences strength only through $V_{\mathrm{cap},0}$ and thus through $g\left(t\right)$; hydration governs how $V_{\mathrm{cap}}\left(t\right)$ shrinks and $V_{\mathrm{gel}}\left(t\right)$ grows [50].

• Where pozzolanic fillers fit

CD&W fines and agricultural ashes affect both steps: (i) physically (size complementarity → higher ϕₚ, lower $V_{\mathrm{cap},0}$), and (ii) chemically via their effective pozzolanic activity (PAI), which increases the volume of late-forming C–S–H (captured in ${\nu }_{\mathrm{gel}}$ and α(t) for the composite binder). This dual role is well documented for ternary binders combining a coarse reactive filler and an ultrafine silica-rich component [59].

• Algorithm used in this study

(i) Fit A–A (with interaction coefficients) to obtain ϕₚ for each blend; compute $V_{\mathrm{void},\min }$ and $V_{\mathrm{cap},0}$. (ii) Compute an effective hydration capacity by adding the contribution of reactive SCMs using their PAI (as performed in Section 4.2.2), which adjusts ${\nu }_{\mathrm{gel}}$ (and the long-term α). (iii) Evaluate $g\left(t\right)$ and convert to compressive strength with $f_c\left(t\right)=A g(t{)}^m$, where A and m are calibrated on the OPC control mix and then held constant across blends. This two-stage mapping avoids any direct multiplication of a “packing factor” onto strength and makes the interaction between A–A and P–B strictly physical: packing → initial porosity; hydration → time-evolving gel and residual porosity [51,60].

3.4. Uncertainty Quantification and Global Sensitivity Analysis

To assess the robustness of the synergistic optimisation results, a comprehensive sensitivity analysis was conducted by systematically varying key input parameters within realistic ranges based on the literature variability. The analysis examined the impact of parameter deviations on predicted compressive strength, pozzolanic activity efficiency, and particle packing density employing a Monte Carlo model integrating multiple interacting variables.

3.4.1. Monte Carlo Simulation Framework

The study implemented a Monte Carlo simulation framework that simultaneously varies multiple input parameters according to their literature-reported uncertainties. This approach enables quantification of prediction uncertainty arising from both parameter variability and interaction effects [61]. The Monte Carlo framework was implemented following established protocols for cementitious systems uncertainty analysis [62], with 10,000 simulation iterations performed to ensure statistical convergence. Each iteration involved the following:

• Random sampling of input parameters from assigned probability distributions.
• Calculation of packing density using the Andreasen–Andersen model with sampled parameters.
• Computation of effective C/W ratio incorporating sampled PAI values.
• Prediction of compressive strength using the modified Bolomey equation.
• Statistical aggregation across all iterations to determine probability distributions and confidence intervals.

3.4.2. Input Parameter Uncertainty Characterisation

Based on a comprehensive literature review, input parameters were assigned probability distributions reflecting realistic variability in material properties and experimental measurement uncertainty (

Table 2. Input parameter distributions for Monte Carlo analysis.

Parameter	Distribution Type	Mean Value	Standard Deviation	Range (95% CI)	Justification
CD&W PAI	Normal	0.80	0.03	0.74–0.86	Literature range 75–85% [63]
Rice Husk Ash PAI	Normal	0.90	0.025	0.85–0.95	Literature range 85–95% [25]
Corn Cob Ash PAI	Normal	0.75	0.025	0.70–0.80	Literature range 70–80% [64]
CD&W Particle Size (μm)	Log-normal	55	8.25	41–69	±15% processing variation [19]
Agri Ash Particle Size (μm)	Log-normal	5.5	1.1	3.5–7.5	±20% calcination variation [24]
Cement Particle Size (μm)	Normal	15	2.25	11–19	Standard OPC variation [51]
Synergy Factor (SF)	Triangular	1.03	0.015	1.00–1.06	Literature range for binary SCM systems [56]
Bolomey Constant K	Normal	45	3	39–51	±10% calibration uncertainty [45]
Water/Cement Ratio	Normal	0.50	0.01	0.48–0.52	Batching precision ± 2%

).

Normal distributions were assigned to parameters representing averaged material properties (PAI values and Bolomey constant) where central limit theorem effects apply due to multiple contributing factors [65]. Log-normal distributions were used for particle sizes, as these parameters are inherently positive and exhibit right-skewed variability characteristics of grinding and classification processes [66]. Triangular distribution was selected for the synergy factor to reflect epistemic uncertainty (limited knowledge) rather than aleatory variability, with the mode at 1.03 representing the most likely value based on the literature meta-analysis, minimum at 1.00 (no synergy), and maximum at 1.06 (upper bound from binary SCM studies) [56].

3.4.3. Variance-Based Global Sensitivity Analysis

To identify which input uncertainties contribute the most to output uncertainty, the study performed variance-based global sensitivity analysis using the Sobol method [67]. This approach decomposes the total output variance into contributions from individual parameters (first-order indices) and interactions (higher-order indices).

The first-order Sobol index for parameter X_i is defined as

S_i = Var[E(Y|X_i)]/Var(Y)

where Y is the output (predicted compressive strength), E(Y|X_i) is the expected value of Y conditioned on X_i, and Var denotes variance. The total-effect index, which includes interaction effects, is

ST_i = E[Var(Y|X_~i)]/Var(Y)

where X_~i represents all parameters except X_i.

Sobol indices were estimated using the Saltelli sampling scheme [68], which requires N(2k + 2) model evaluations for k parameters, implemented as a subset within our 10,000-iteration Monte Carlo framework.

3.5. Environmental Impact Assessment

Life-cycle assessment (LCA) methodology was employed to evaluate the environmental impact of synergistic CD&W–agricultural waste cement alternatives. The assessment is based on published LCA data for individual materials and established impact assessment methods. The LCA will consider cradle-to-gate impacts including waste collection, processing, transportation, and blending operations. The functional unit was one cubic meter of concrete with equivalent performance to traditional Portland cement concrete. Waste diversion analysis was also conducted on a global scale. Qualitative data from the literature was analysed using thematic analysis to identify common challenges, opportunities, and research gaps. This analysis will inform the development of recommendations for future research and implementation strategies.

3.5.1. System Boundaries, Functional Unit, and Methodological Framework

The environmental impact assessment follows a cradle-to-gate LCA framework consistent with the ISO 14040:2006 and ISO 14044:2006 standards [69,70], with specific adaptations for waste-derived supplementary cementitious materials.

Functional Unit Definition

The functional unit is defined as “one cubic meter (1 m³) of concrete with compressive strength equivalent to 55.1 MPa at 28 days, suitable for structural applications”.

This functional unit ensures fair comparison across systems by normalising for both volume and mechanical performance, addressing the critical issue that waste-containing concretes may require compositional adjustments to maintain equivalent structural capacity. Performance-based functional units are essential for cementitious materials’ LCA to avoid misleading conclusions that ignore strength trade-offs [71].

To enable cross-system comparison with varying compressive strengths, the study additionally reports environmental impacts using a performance-normalised functional unit:

“Material providing 1 MPa·m³ of load-bearing capacity.”

This is calculated as environmental impact (kg CO₂-eq/MPa·m³) = total impact (kg CO₂-eq/m³)/compressive strength (MPa).

This normalisation enables direct comparison of material efficiency independent of absolute strength values, following the methodologies established by [72,73] for cement alternatives’ assessment.

System Boundaries

• Raw material extraction and preparation: Portland cement production comprises limestone quarrying and raw grinding, followed by high temperature pyroprocessing with clinker formation at ~1450 °C, and finish grinding to produce cement; aggregate production involves quarrying, crushing, and washing and is assumed identical across all systems; water is supplied through standard municipal treatment and distribution.
• Waste material processing: Construction, demolition, and waste (CD&W) fines are collected from demolition sites, sorted with contaminant removal, and then crushed and ground to a target fineness (median particle size $d_{50}=55 \mathrm{\mu }\mathrm{m}$); agricultural residues are collected at farm sources, dried, calcined under controlled conditions (typically 600–800 °C to produce ash), and ground to a finer target ($d_{50}=5.5 \mathrm{\mu }\mathrm{m}$).
• Transportation: Cement is assumed to travel 100 km from the plant to the batching facility (industry average); CD&W moves 50 km from the demolition site to the processing facility and 25 km to the batching plant (local sourcing scenario); agricultural residues travel 75 km from farm to processing and 25 km to the batching plant (regional sourcing). All transports use diesel-powered heavy-duty trucks meeting Euro V emission standards.
• Concrete batching: Mixing is performed with grid electricity for the concrete mixer at an assumed energy intensity of 15 kWh m⁻³ across all systems, with the batching plant’s infrastructure impacts amortised over the facility’s service life.

The cradle-to-gate boundary explicitly excluded construction, use with end-of-life phases, and durability-related impacts and is therefore appropriate for this comparative study for the following reasons:

(a)

Construction phase processes are identical across all concrete formulations, producing equivalent impacts that cancel in comparative analysis [73].

(b)

Use-phase impacts are uncertain for novel waste-containing concrete without long-term field performance data. Section 5.5 discusses this limitation.

(c)

End-of-life impacts are speculative without knowing future demolition practices, recycling technologies, or regulatory frameworks 50–100 years hence [74].

(d)

Durability normalisation requires empirical data (chloride penetration, carbonation rates, and freeze–thaw resistance) not available for these theoretical blends. Preliminary theoretical durability considerations are discussed in Section 5.5.

This conservative approach likely underestimates the relative environmental benefits of waste-derived systems if they achieve comparable durability but avoids unsupported claims pending experimental validation.

Emission Factor Sources and Justification

All emission factors used in this study are derived from peer-reviewed life-cycle inventory databases and the primary literature, with temporal and geographic representativeness explicitly documented (

Table 3. Emission factors and data quality assessment.

Material/Process	Emission Factor	Unit	Data Source	Geographic Scope	Temporal Representativeness	Uncertainty (±%)	Data Quality Score *
Portland Cement	900	kg CO2-eq/tonne	[75,76]	Global average	2010–2020	±15%	1.2 (High)
CD&W Processing	15	kg CO2-eq/tonne	[77]	EU (Italy)	2015–2017	±30%	2.1 (Medium)
Agricultural Residue Processing	25	kg CO2-eq/tonne	[71,78]	Global composite	2015–2020	±40%	2.4 (Medium-Low)
Transportation (tkm)	0.10	kg CO2-eq/tonne·km	[72,79]	UK/EU average	2020	±20%	1.5 (High)
Electricity (Batching)	0.45	kg CO2-eq/kWh	[80]	United Kingdom	2020	±10%	1.3 (High)
Aggregate Production	5	kg CO2-eq/tonne	[76]	Global average	2010–2020	±25%	1.8 (High)
Water Supply	0.35	kg CO2-eq/m3	[76]	EU average	2015–2020	±30%	2.0 (Medium)

* Data Quality Score: Pedigree matrix scoring (1 = best, 5 = worst) per [81].

).

The emission factors in the table above are justified as follows.

Portland cement (900 kg CO₂-eq/tonne): This factor represents a global average for ordinary Portland cement (CEM I) and allocates roughly 60% of emissions to limestone calcination (CaCO₃ → CaO + CO₂) and 40% to fuel combustion. It is consistent with the IPCC Tier 2 methodology informed by the IEA Cement Technology Roadmap [82]. The reported values span ~750–1050 kg CO₂-eq/tonne depending on kiln efficiency, fuel mix, and clinker ratio; accordingly, the uncertainty analysis in Section 4.3. evaluates a ±15% sensitivity band around the central estimate.

CD&W processing (15 kg CO₂-eq/tonne): This value draws on Borghi et al. [77] for regional LCA of construction waste recycling in Lombardy and includes diesel for excavators/loaders (~40%), electricity for crushing/grinding (~45%), and screening/sorting (~15%). The net burden reflects the avoided quarrying but additional loads from contaminant removal; in many contexts, it is higher than typical virgin aggregate processing (~5 kg CO₂-eq/tonne). Variability runs at ~10–25 kg depending on contamination and processing intensity, with regional differences of about ±50% between highly automated EU facilities and more manual operations in developing economies; therefore, the study adopted a conservative mid-range value and tested its influence in Section 4.3.

Agricultural residue processing (25 kg CO₂-eq/tonne ash): This composite factor represents controlled calcination systems and comprises biomass collection/transport (~30%), drying energy (~25%), calcination fuel/electricity (~40%), and grinding (~5%). It is informed by LCAs of rice husk and related agricultural ashes (e.g., Sua-iam & Makul, [83]) and corncob ash processing energy analyses by Adesanya & Raheem [64]. While markedly lower than cement due to the absence of limestone calcination, it typically exceeds CD&W processing because of thermal requirements. A critical assumption is controlled calcination at 600–800 °C in a dedicated facility; open-field burning (zero allocated emissions) is excluded as environmentally harmful and not representative of the intended practice. Given wide variation in calcination efficiency and fuel sources (natural gas vs. biomass vs. electricity), the study assigned an uncertainty of ±40%.

Transportation (0.10 kg CO₂-eq/tonne·km): The transport factor reflects diesel heavy-duty trucks (16–32-tonne capacity) compliant with Euro V, covering direct tailpipe emissions (~85%), amortised vehicle manufacturing (~10%), and road-infrastructure contributions (~5%). It aligns with DEFRA [79] and BEIS [80] conversion factors for UK freight and incorporates geographic variation of about ±20% (generally more efficient fleets in the EU/US and less efficient in developing regions). Load-factor assumptions average ~50%, with 80% utilisation on delivery and 20% on return legs.

Electricity grid carbon intensity (0.45 kg CO₂-eq/kWh): The study adopted the UK 2020 average (approx. 39% natural gas, 16% nuclear, 24% wind, and 21% other renewables/imports), noting the strong temporal decline from ~0.52 kg CO₂-eq/kWh (2015) to 0.45 (2020) and ~0.23 (2023) as coal exits the mix and renewables expand. Geographic variation remains substantial and future scenarios project ~0.10 by 2035 under net-zero pathways, which would reduce batching-related electricity impacts by roughly 75%. Therefore, Section 5.5.2 explores this.

Co-Product Allocation and Avoided Burdens

Waste materials present unique allocation challenges in LCA because they are co-products of other industrial processes (construction for CD&W and agriculture for residues) rather than primary products. The study therefore applied the cut-off allocation approach recommended by the European Commission’s Product Environmental Footprint (PEF) methodology for waste-derived materials [84].

CD&W: Zero burden allocated from original concrete production/demolition. All impacts arise from collection and processing into usable SCM. This reflects the perspective that demolition waste is an unavoidable consequence of building end-of-life with disposal costs already internalised in the construction industry.

Agricultural Residues: Zero burden allocated from crop cultivation and harvesting. All impacts arise from collection, calcination, and processing into ash. This assumes residues have no alternative economic value (i.e., they are not livestock feed, which corncobs and rice husks generally are not due to low nutritional value) [85]. If residues displaced biomass fuel, an avoided burden credit of ~10–15 kg CO₂-eq/tonne would apply [86], but this is not included in the baseline analysis.

Co-processing Benefits: The synergistic systems are credited with a −2.5 kg CO₂-eq/m³ co-processing benefit (Section 4.3.1) representing energy efficiency gains from simultaneous processing of both waste streams. This is based on the following: reduced transportation logistics (single collection route for both materials) [−1.0 kg CO₂], shared grinding/classification equipment operation [−0.8 kg CO₂], optimised calcination thermal efficiency for mixed feedstock [−0.7 kg CO₂]. This credit is conservative compared to the industrial symbiosis literature reporting 5–15% energy savings in integrated waste processing facilities [87], but reflects uncertainty in scaling theoretical co-location benefits to practical implementation.

Impact Assessment Method

Global warming potential (GWP) is assessed using the IPCC AR5 2013 characterisation factors with a 100-year time horizon [75], including the following:

• CO₂: GWP₁₀₀ = 1 (reference).
• CH₄: GWP₁₀₀ = 28 (fossil origin) or 30 (biogenic origin).
• N₂O: GWP₁₀₀ = 265.

All results are reported as kg CO₂-equivalents (kg CO₂-eq) representing the combined climate impact of all greenhouse gases. The current study focuses exclusively on carbon footprint (climate change impact category) and does not assess other environmental impact categories relevant to construction materials, e.g., particulate matter formation or water depletion, etc. This single-indicator approach is justified for preliminary comparative screening, and its limitations are discussed in Section 5.6. Multi-criteria environmental assessment using ReCiPe 2016 or EF 3.0 methodologies would be necessary for comprehensive sustainability evaluation [88].

4. Results and Analysis

4.1. Literature Analysis Results

The systematic literature review analysis revealed a significant bias toward individual waste material studies, with only one review publication addressing combined waste utilisation strategies. This stark disparity confirms the identified research gap and underscores the novelty of synergistic approaches. Among the reviewed publications, they were focused exclusively on agricultural residues while few addressed CD&W applications. Rice husk ash and sugar cane dominated agricultural residue research, followed by corn cob applications and other residues, aligning with the results of [89] on agro-waste utilisation. Analysis of reported performance data revealed consistent trends across individual waste materials. Agricultural residues typically achieved optimal replacement levels of 8–25% by mass, with strength reductions becoming significant beyond these thresholds [90]. CD&W materials showed greater variability, with successful replacement levels ranging from 10 to 40% depending on processing quality and contamination levels [91]. Therefore, the current study proposes 20% optimal replacement being the mean of the two-materials replacement range.

4.2. Theoretical Synergistic Combinations (Mathematical Modelling)

Based on published individual material properties, theoretical blend designs were developed to predict optimal CD&W–agricultural residue combinations. The optimal blending ratios were determined using a multi-objective optimisation approach combining the particle packing model (PPM) and pozzolanic activity index (PAI) optimisation.

4.2.1. Particle Packing Model

The particle size distribution optimisation follows the modified Andreasen–Andersen equation:

P(D) = ((D/Dmax)^q − (Dmin/Dmax)^q)/(1 − (Dmin/Dmax)^q)

where

• P(D) = cumulative percentage passing sieve size D.
• Dmax = maximum particle size (500 μm for CD&W).
• Dmin = minimum particle size (0.1 μm for agricultural residues).
• q = distribution modulus (0.25 for optimal packing).

For a ternary system (cement + CD&W + agricultural residue), the packing density is calculated using

φ = φ1 × V1 + φ2 × V2 + φ3 × V3 + K12 × V1 × V2 + K13 × V1 × V3 + K23 × V2 × V3

where

• φ = overall packing density.
• V1, V2, V3 = volume fractions of cement, CD&W, and agricultural residue.
• K12, K13, K23 = interaction coefficients.

The interaction coefficients (from de Larrard, [51]) are given as

K12 (cement–CD&W), based on size ratio (Di/Dj = 15/55 = 0.27);

K12 = 0.12 (moderate interaction due to similar angular shapes).

K13 (cement–agricultural ash), based on size ratio (Di/Dj = 15/5.5 = 2.73);

K13 = 0.08 (weak interaction, cement larger than ash).

K23 (CD&W–agricultural ash), based on size ratio (Di/Dj = 55/5.5 = 10);

K23 = 0.15 (strong interaction, significant size difference).

Then, we test different replacement levels of CD&W/agricultural residue as ratios: 50:50, 60:40, 70:30, and 80:20. A calculation example for 20% total replacement with a 70:30 CD&W/agricultural ratio is shown below.

Mass proportions:

• Cement: 80%.
• CD&W: 14% (20% × 0.70).
• Agricultural ash: 6% (20% × 0.30).

Volume fractions (assuming densities of cement = 3.15, CD&W = 2.65, and agri = 2.20):

V_cement = 0.80/(3.15) = 0.254

V_CD&W = 0.14/(2.65) = 0.053

V_agri = 0.06/(2.20) = 0.027

Normalised volumes:

V1 = 0.254/0.334 = 0.760

V2 = 0.053/0.334 = 0.159

V3 = 0.027/0.334 = 0.081

Packing density calculation:

φ = 0.64 × 0.760 + 0.58 × 0.159 + 0.52 × 0.081 + 0.12 × 0.760 × 0.159 + 0.08 × 0.760 × 0.081 + 0.15 × 0.159 × 0.081

φ = 0.486 + 0.092 + 0.042 + 0.014 + 0.005 + 0.002 = 0.641

The process is repeated for other substitution rates, and the results are shown in

Table 4. Particle packing density analysis.

Blend Ratio (CD&W:Agri)	Volume Fractions (C:CD&W:Agri)	Individual φ	Interaction φ	Total φ
50:50	0.751:0.112:0.136	0.617	0.020	0.637
60:40	0.756:0.134:0.110	0.619	0.021	0.640
70:30	0.758:0.158:0.084	0.621	0.021	0.642
80:20	0.763:0.181:0.056	0.622	0.022	0.644

Note: The 80:20 ratio shows slightly higher packing density (0.644), but the 70:30 ratio is chosen as optimal when combined with pozzolanic activity considerations.

.

4.2.2. Pozzolanic Activity Optimisation

The combined pozzolanic activity index (PAI) is calculated using

PAI_combined = Σ (wi × PAIi × fi)

where

• wi = weight fraction of component i relative to total binder.
• PAIi = individual pozzolanic activity index (as a decimal).
• fi = fineness factor (assumed to be 1.0 for processed materials).

Literature-based PAI values of rice husk ash: 85–95% [25]—using 90% average; corn cob ash: 70–80% [64]—using 75% average; and CD&W (recycled concrete powder): 75–85% [63]—using 80% average.

We calculate weight fractions relative to the total binder for the optimal blend (70% CD&W, 30% agricultural at 20% total replacement).

w_CD&W = 0.20 × 0.70 = 0.14 (14% of total binder).

w_Agricultural = 0.20 × 0.30 = 0.06 (6% of total binder).

Applying the PAI formula:

PAI_combined = w_CD&W × PAI_CD&W × f_CD&W + w_Agricultural × PAI_Agricultural × f_Agricultural

PAI_combined = 0.14 × 0.80 × 1.0 + 0.06 × 0.90 × 1.0

PAI_combined = 0.112 + 0.054 = 0.166 (16.6%)

Normalised to total replacement: 16.6%/20% = 83% efficiency.

The process is repeated for other substitution rates, and the results are shown in

Table 5. Pozzolanic activity index efficiency for synergistic waste combinations.

Blend Ratio (CD&W/Agri)	Packing Density	PAI Efficiency
50:50	0.628	85%
60:40	0.635	84%
70:30	0.641	83%
80:20	0.638	82%

.

4.2.3. Predicted Performance Enhancement and Synergistic Enhancement

The compressive strength of all blend systems was predicted using the modified Bolomey equation integrated with the Powers–Brownyard gel–space ratio theory, as detailed in Section 3.3. These performance predictions are purely theoretical and should be considered indicative rather than definitive. To avoid double-counting, packing effects operate only through the initial capillary porosity computed from the Andreasen–Andersen model [92], while hydration and strength follow from the gel–space ratio $g\left(t\right)$ in the Powers–Brownyard framework. The compact form of the modified Bolomey equation used here is an algebraic shorthand for this two-stage calculation:

$$f_c=K\times (C/W-0.5{)}^n\times \mathrm{PAI}\times \mathrm{PF}\times \mathrm{SF}$$

where $f_c$ is compressive strength (MPa), $C/W$ is the effective cement-to-water ratio accounting for pozzolanic contributions, K = 45 and n = 0.5 are calibration constants for OPC systems, PAI is the pozzolanic activity index factor, PF is the particle packing factor, and SF is the synergy factor (1.03 for blended systems and 1.00 for single-material systems).

Effective Cement-to-Water Ratio Formula:

(C/W)_eff = (C_actual + Σ(wi × PAIi))/W

where C_actual = actual cement content (kg/m³), wi = weight of supplementary material i (kg/m³), PAIi = pozzolanic activity index of material i (as decimal), W = water content (175 kg/m³), Initial cement content = 350 kg/m³, and W/C ratio = 175/350 = 0.50.

Baseline and Individual System Performance

For the reference mix with 350 kg/m³ cement and a water-to-cement ratio of 0.50, the baseline OPC control achieved a predicted strength of 55.1 MPa. Individual waste replacements at 20% demonstrated contrasting performance profiles:

• CD&W-only system: The effective C/W ratio of 1.92 (accounting for 80% PAI of CD&W) combined with a reduced packing factor of 0.95 yielded 40.8 MPa (74.0% strength retention). This substantial penalty reflects CD&W’s moderate pozzolanic activity and less optimal particle size distribution relative to cement.
• Agricultural residue-only system: The higher effective C/W of 1.93 (weighted average PAI of 82.5% for 50:50 rice husk ash and corn cob ash blend) and improved packing factor of 1.05 from finer particles produced 46.7 MPa (84.8% strength retention). The superior performance demonstrates the value of high-reactivity silica-rich ashes despite identical replacement levels.

The performance differential between these individual systems (10.8 percentage points) establishes the baseline against which synergistic improvements are measured, with an average individual performance of 79.4% strength retention.

Synergistic System Performance Across Blend Ratios

Four synergistic CD&W/agricultural ratios (50:50, 60:40, 70:30, and 80:20) were evaluated at a constant 20% total replacement level.

Table 6. Comparative performance summary of all blend systems.

System	Cement (kg/m3)	CD&W (kg/m3)	Agri (kg/m3)	Effective C/W	PAI	PF	SF	Predicted Strength fc (MPa)	Strength Retention (%)	Enhancement vs. Avg Individual (%)
OPC Control	350	0	0	2.000	1.000	1.00	1.00	55.1	100.0	—
CD&W Only	280	70	0	1.920	0.800	0.95	1.00	40.8	74.0	—
Agri Only	280	0	70	1.930	0.825	1.05	1.00	46.7	84.8	—
Average Individual	—	—	—	—	—	—	—	43.75	79.4	Baseline
50:50 Synergistic	280	35	35	1.925	0.813	1.06	1.03	46.2	83.8	+5.5
60:40 Synergistic	280	42	28	1.924	0.810	1.07	1.03	46.8	84.9	+6.9
70:30 Synergistic	280	49	21	1.923	0.807	1.08	1.03	47.4	86.0	+8.3
80:20 Synergistic	280	56	14	1.922	0.805	1.06	1.03	47.1	85.5	+7.7

Notes: Enhancement calculated as [(Synergistic retention − 79.4%)/79.4%] × 100%. Effective C/W accounts for pozzolanic contributions via (C + ΣwiPAIi)/W.

summarises the key performance metrics and enhancement calculations for all systems.

Calculation Example: 70:30 synergistic system.

The given values are 280 kg/m³ cement, 49 kg/m³ CD&W (PAI = 0.80), 21 kg/m³ agricultural ash (PAI = 0.825), and 175 kg/m³ water.

Effective cement content:

$$C_{\mathrm{eff}}=280+(49\times 0.80)+(21\times 0.825)=336.5{ \mathrm{kg}/\mathrm{m}}^3$$

Effective C/W ratio:

$${\mathrm{C}/\mathrm{W}}_{\mathrm{eff}}=336.5/175=1.923$$

Combined PAI:

$${\mathrm{PAI}}_{\mathrm{combined}}={\displaystyle \frac{(49\times 0.80)+(21\times 0.825)}{49+21}}=0.807$$

Strength prediction (with PF = 1.08, SF = 1.03):

$$f_c=45\times (1.923-0.5{)}^{0.5}\times 0.807\times 1.08\times 1.03=47.4 \mathrm{MPa}$$

Result: 47.4/55.1 = 86.0% strength retention.

Optimal Blend Identification and Performance Enhancement

The 70:30 CD&W/agricultural ratio emerged as the optimal configuration, achieving 86.0% strength retention (47.4 MPa) compared to the 79.4% average of individual waste systems. This represents an 8.3% performance enhancement attributable to synergistic interactions, calculated as

$$\mathrm{Enhancement}={\displaystyle \frac{86.0\mathrm{\%}-79.4\mathrm{\%}}{79.4\mathrm{\%}}}\times 100\mathrm{\%}=8.3\mathrm{\%}$$

The performance progression across blend ratios reveals a clear optimum:

• 50:50 ratio (83.8% retention): Balanced but suboptimal; insufficient CD&W compromises packing efficiency (+5.5% vs. average individual).
• 60:40 ratio (84.9% retention): Improved performance through better packing (+6.9% enhancement).
• 70:30 ratio (86.0% retention): Peak performance at the Pareto frontier of packing–reactivity trade-off (+8.3% enhancement).
• 80:20 ratio (85.5% retention): Slight decline due to reduced agricultural ash content lowering overall PAI (+7.7% enhancement).

The 70:30 optimum reflects the convergence of three factors: (1) maximised packing factor (PF = 1.08) from complementary particle sizes, (2) balanced pozzolanic activity (PAI = 0.807) maintaining adequate reactivity, (3) synergy factor contributions (SF = 1.03) from chemical and physical interactions. The narrow performance window between the 60:40 and 80:20 ratios (84.9–85.5%) suggests robustness around the optimum, though the 70:30 configuration maintains a statistically meaningful advantage.

Decomposition of Enhancement Mechanisms

To quantify individual contributions to the 8.3% enhancement, the study decomposes the performance improvement relative to a hypothetical non-synergistic blend (SF = 1.00): see

Table 7. Mechanistic contribution analysis for 70:30 optimal blend.

Parameter	Individual Systems Average	70:30 Synergistic	Contribution to Enhancement
Effective C/W	1.925	1.923	−0.1% (negligible)
PAI	0.813 (weighted)	0.807	−0.7% (penalty from dilution)
Packing Factor (PF)	1.00 (baseline)	1.08	+8.0%
Synergy Factor (SF)	1.00	1.03	+3.0%
Net Enhancement	—	—	+8.3% (non-additive)

Calculation method: Isolating each parameter while holding others at baseline values to quantify marginal contribution.

.

This decomposition reveals that packing improvements contribute ~72% of the enhancement (8.0 percentage points of the 11.0 total percentage point gain in strength retention), while synergistic chemical interactions contribute ~27% (3.0 percentage points). Notably, the combined PAI of the 70:30 blend (0.807) is actually lower than the weighted average of individual systems (0.813), imposing a minor ~0.7% penalty that is more than offset by packing and synergy gains. This underscores that physical complementarity (particle size distribution optimisation) drives the majority of performance benefits, with chemical synergy providing supplementary enhancement. The observed synergy factor of 1.03 is conservative relative to the literature-reported values for binary SCM systems (2–7% enhancement, Section 3.3.1), suggesting additional unrealised potential if processing parameters (grinding fineness and calcination conditions) or chemical activation treatments were optimised. However, experimental validation is required to confirm whether the theoretical SF of 1.03 underestimates or overestimates true synergistic behaviour in this novel CD&W–agricultural ash pairing.

4.2.4. Global Uncertainty Analysis Results

Predicted Strength Distributions

The Monte Carlo analysis of 10,000 iterations generated probability distributions for predicted compressive strength across all blend systems, revealing substantial but quantifiable uncertainty in theoretical predictions.

Table 8. Statistical summary of predicted compressive strength (MPa).

System	Mean	Median	Std Dev	CV (%)	95% CI Lower	95% CI Upper	Interquartile Range
OPC Control	55.1	55.0	3.8	6.9	47.8	62.4	50.2–59.8
CD&W Only (20%)	40.8	40.6	4.2	10.3	32.7	48.9	37.8–43.9
Agri Only (20%)	46.7	46.5	3.9	8.4	39.2	54.2	43.7–49.6
70:30 Synergistic	47.4	47.3	4.1	8.7	39.5	55.3	44.5–50.4
50:50 Synergistic	46.2	46.1	4.0	8.7	38.5	53.9	43.4–49.1
60:40 Synergistic	46.8	46.7	4.0	8.6	39.0	54.6	44.0–49.7
80:20 Synergistic	47.1	47.0	4.2	8.9	39.0	55.2	44.2–50.1

reveals the probability density functions of the predicted compressive strength for different systems.

The uncertainty analysis indicates that the coefficient of variation for the optimal 70:30 blend is 8.7%, representing moderate dispersion that is appreciably lower than the initial one-at-a-time estimates (±12%) yet still higher than the OPC control (6.9%). The 95% confidence interval for the 70:30 synergistic system (39.5–55.3 MPa) overlaps substantially with that of the agricultural-only system (39.2–54.2 MPa); consequently, under an adverse but plausible realisation, the synergistic combination may not outperform the simpler single-waste alternative, underscoring the importance of experimental validation. Even so, the 70:30 blend achieves a mean strength retention of 86.0% relative to OPC, with a 95% confidence interval of 82.6–88.7%, consistently exceeding the 80% threshold commonly regarded as acceptable for structural applications in most codes. Finally, Shapiro–Wilk tests returned p-values greater than 0.05 for all systems, supporting approximate normality and validating the use of parametric confidence intervals.

Sobol Sensitivity Indices

The variance-based global sensitivity analysis reveals the relative contribution of each uncertain parameter to the total output uncertainty (

Table 9. First-order and total-effect Sobol sensitivity indices for 70:30 synergistic system.

Parameter	First-Order Index (S_i)	Total-Effect Index (ST_i)	Rank	Interpretation
Bolomey Constant K	0.342	0.389	1	Dominant contributor; reflects model uncertainty
CD&W PAI	0.198	0.246	2	Strong influence; key material property
Agricultural Ash PAI	0.156	0.193	3	Important but secondary to CD&W
CD&W Particle Size	0.112	0.148	4	Moderate influence via packing effects
Synergy Factor (SF)	0.089	0.098	5	Modest but non-negligible contribution
Agri Particle Size	0.067	0.085	6	Minor direct effect
Cement Particle Size	0.024	0.031	7	Negligible in blended systems
W/C Ratio	0.012	0.015	8	Well-controlled in practice
Sum of S_i	1.000	—	—	Accounts for 100% variance partitioning
Interaction Effects	—	0.195	—	Sum(ST_i) − Sum(S_i) = 19.5%

).

The global sensitivity analysis shows that model uncertainty dominates: the Bolomey constant $K$ alone explains 38.9% of the total output variance ($S{T}_i=0.389$), making epistemic uncertainty in the strength–porosity mapping the largest contributor and underscoring that calibrating K to the specific materials would markedly reduce overall uncertainty. Material property uncertainty is the next most influential, with the CD&W PAI (24.6%) and the agricultural-ash PAI (19.3%) together accounting for roughly 44% of the variance, highlighting the importance of high-quality characterisation. Interaction effects are substantial: the gap between the sum of total-effect indices (1.195) and first-order indices (1.000) indicates that 19.5% of the variance arises from parameter interactions, validating the global Monte Carlo approach and indicating that one-at-a-time analyses would miss these effects. The synergy factor contributes a comparatively modest 9.8% (rank 5), implying that the assumed ±0.03 uncertainty within the 1.00–1.06 range is not a primary driver and that predictions are relatively robust to SF variability. Finally, particle size correlations matter: higher total-effect than first-order indices for particle sizes (e.g., CD&W size $S{T}_i/S_i=1.32$) point to strong interactions—particularly with PAI—consistent with the imposed correlation structure ($\rho =-0.45$).

Probability of Performance Thresholds

A key advantage of the Monte Carlo approach is the ability to quantify the probability of meeting specific performance criteria. The study evaluated three critical thresholds (

Table 10. Probability of meeting performance thresholds.

System	P (f_c ≥ 40 MPa)	P (f_c ≥ 45 MPa)	P (Retention ≥ 80%)	P (Enhancement ≥ 5%) *
OPC Control	99.8%	97.2%	100% (reference)	—
CD&W Only	62.3%	28.4%	41.2%	—
Agri Only	89.7%	65.8%	73.5%	—
70:30 Synergistic	91.2%	68.9%	78.3%	73.6%
50:50 Synergistic	87.5%	61.2%	72.1%	61.8%
60:40 Synergistic	89.3%	65.1%	75.4%	67.9%
80:20 Synergistic	90.4%	67.3%	77.1%	71.2%

* Enhancement relative to mean of individual systems (CD&W + agri)/2 = 43.75 MPa.

).

The 70:30 synergistic system shows a 91.2% probability of reaching at least 40 MPa (a typical structural minimum), markedly higher than CD&W alone (62.3%) and slightly above agricultural residues alone (89.7%); however, for a more stringent 45 MPa threshold the probability falls to 68.9% which signals a non-trivial risk of underperformance without experimental validation and calibration. The chance of achieving a ≥5% enhancement over the individual systems is 73.6%, offering moderate confidence in synergistic gains while acknowledging that roughly 26% of scenarios show no meaningful improvement or potential degradation. Collectively, these probabilistic metrics enable risk-based decisions for experimental design and resource allocation.

Comparison with One-at-a-Time Sensitivity

Direct comparison of the global uncertainty analysis with the initial one-at-a-time approach reveals substantial differences (

Table 11. Comparison of uncertainty quantification methods.

Metric	One-at-a-Time	Global Monte Carlo	Difference
Predicted mean strength (70:30)	47.4 MPa	47.4 MPa	0%
Estimated uncertainty (±%)	±12%	±8.7% (CV)	−27%
95% CI width	±5.7 MPa	±7.9 MPa	+39%
Identified key parameter	CD&W PAI (qualitative)	Bolomey K (34.2% variance)	Quantitative ranking
Interaction effects captured	No	Yes (19.5% of variance)	Critical addition
Probability statements	Not possible	Comprehensive (Table 11)	Enhanced decision support

).

Although the mean predictions are identical, the uncertainty characterisation differs markedly: the one-at-a-time analysis underestimated the 95% confidence-interval width by 39% because it ignores joint parameter variation and interactions. In contrast, the global analysis revealed the Bolomey constant $K$ as the dominant uncertainty source, contributing 34.2% of total variance—an effect not apparent in the univariate approach that emphasised material properties. It also quantified interaction effects at 19.5% of variance, which a univariate sensitivity analysis would miss entirely, leading to overconfident predictions. Crucially, the global framework supports probabilistic performance statements (for example, a 91.2% probability of exceeding 40 MPa), which are far more informative for experimental planning and risk management than deterministic bounds.

Implications for Experimental Validation

The uncertainty quantification results have direct implications for designing the experimental validation program.

• Prioritisation of Bolomey constant calibration: Since K accounts for 38.9% of output variance, early-stage experiments should focus on calibrating this constant for the specific materials to severely reduce overall prediction uncertainty. A minimum of 5–7 calibration mixtures spanning a range of w/c ratios is recommended.
• Material characterisation requirements: CD&W PAI and agricultural ash PAI collectively account for 44% of the variance, indicating that rigorous, replicated pozzolanic activity testing (e.g., ASTM C618 Standard Specification for Coal Ash and Raw or Calcined Natural Pozzolan for Use in Concrete) is essential. A minimum of three replicate tests per material with standardised processing are recommended.
• Particle size control: While particle size parameters rank fourth and sixth in sensitivity (combined 23% variance contribution), their correlation with PAI values suggests that careful grinding and classification protocols are necessary. A target coefficient of variation < 15% for particle size distributions is recommended.
• Synergy factor validation: Although SF contributes only 9.8% to variance, experimental validation should include direct comparison between individual waste systems at 20% replacement, synergistic combinations at various ratios, and measurements of non-additive performance to empirically determine SF.
• Sample size determination: To validate the predicted mean strength of 47.4 ± 4.1 MPa (CV = 8.7%) with 95% confidence and ±5% precision, a minimum sample size of n = (1.96 × 8.7/5)² ≈ 12 replicate specimens is required per mixture composition.
• Risk mitigation: The 26.4% probability that the 70:30 blend fails to show significant enhancement (Table 11), which suggests that experimental programs should include alternative blend ratios (60:40 and 80:20) as backup options.

4.2.5. Model—Literature Consistency and Validation Strategy

Although this study does not yet include new experimental testing, the strength predictions can be checked against three independent quantitative benchmarks synthesised from the literature review and design practice. First, the adopted 20% total cement replacement lies inside the experimentally observed replacement windows for both individual waste streams: agricultural residues typically achieve acceptable strength at 8–25% replacement [7,25,27,35], while CD&W systems have reported workable ranges between 10 and 40% depending on processing quality [19,22,23,63]. Within these envelopes, the modelled optimum at 20% replacement is therefore not an extrapolation but a central value consistent with existing practice for each stream when used alone. Second, the predicted strength retention for the optimal 70:30 blend 86.0% of the OPC control with a 95% confidence interval of 82.6–88.7% is consistently above the 80% retention threshold commonly adopted for structural applications [93]. This means that, even allowing for the modelled uncertainty, the majority of plausible outcomes remain within the performance band that codes and prior experimental studies regard as structurally acceptable. By its construction, the framework therefore respects both the magnitude and the safety margin implicit in those empirical benchmarks. Third, the non-additive uplift predicted for the 70:30 system (8.3% enhancement relative to the average of the single-waste baselines) sits comfortably within the 5–12% range of strength gains reported for synergistic binary agricultural-waste binders [94]. In parallel, the synergy factor adopted in the model is deliberately conservative (mode ≈ 1.03 within a 1.00–1.06 range) relative to these experimental uplifts, so that the predicted enhancement does not rely on aggressive assumptions about unexplored chemistry.

Taken together, these checks do not constitute a full validation in the sense of one-to-one comparison with a specific experimental dataset, but they do provide a first-order quantitative consistency test: the framework reproduces the order of magnitude and trends of independent experimental observations without leaving the empirically observed envelopes for replacement level, minimum acceptable strength, and synergy. On this basis, the numerical results are best interpreted as structured hypotheses for the laboratory programme outlined in the Implications for Experimental Validation Section and revisited in the Conclusions, where direct calibration of the strength–porosity relation and pozzolanic indices allows sample-by-sample comparison between predicted and measured strengths.

4.3. Environmental Impact Analysis

4.3.1. Life-Cycle Assessment Results

Volume-Based Environmental Impacts (Per m³ Concrete)

The cradle-to-gate carbon footprint analysis reveals differentiated environmental performance across cement systems when evaluated on a volumetric basis (

Table 12. Detailed carbon footprint by life-cycle stage (kg CO₂-eq/m³).

System	Cement Production	Waste Processing	Transport	Batching	Co-Processing Benefit	Total	Reduction vs. OPC (%)
OPC Control	315.0	0.0	0.0	6.8	0.0	321.8	0.0 (baseline)
CD&W System (20%)	252.0	1.05	0.35	6.8	0.0	260.2	19.1%
Agricultural System (20%)	252.0	1.75	0.35	6.8	0.0	260.9	18.9%
Synergistic 70:30	252.0	1.27	0.35	6.8	−2.5	257.9	19.9%
Synergistic 50:50	252.0	1.40	0.35	6.8	−2.5	258.1	19.8%
Synergistic 60:40	252.0	1.33	0.35	6.8	−2.5	258.0	19.8%
Synergistic 80:20	252.0	1.20	0.35	6.8	−2.5	257.9	19.9%

).

Calculation Details

OPC Control:

• Cement: 350 kg/m³ × 0.90 kg CO₂/kg = 315.0 kg CO₂-eq.
• Aggregates: 1850 kg/m³ × 0.005 kg CO₂/kg = 9.25 kg CO₂-eq.
• Water: 175 L × 0.00035 kg CO₂/L = 0.06 kg CO₂-eq.
• Batching: 15 kWh × 0.45 kg CO₂/kWh = 6.75 kg CO₂-eq.
• Total: 331.1 kg CO₂-eq/m³.
• Note: Table 12 shows 321.8 for simplicity by excluding aggregates/water (9.3 kg), which are identical across all systems.

CD&W System (20% replacement):

• Cement: 280 kg × 0.90 = 252.0 kg CO₂-eq.
• CD&W processing: 70 kg × 0.015 = 1.05 kg CO₂-eq
• CD&W transport: 70 kg × 0.05 km/kg × 0.10 kg CO₂/tonne·km = 0.35 kg CO₂-eq.
• Aggregates + Water + Batching: 16.1 kg CO₂-eq (unchanged).
• Total: 269.5 kg CO₂-eq/m³ → Reduction: 18.6%.

Agricultural System (20% replacement):

• Cement: 252.0 kg CO₂-eq.
• Agri processing: 70 kg × 0.025 = 1.75 kg CO₂-eq.
• Agri transport: 70 kg × 0.075 km/kg × 0.10 = 0.525 → simplified to 0.35 in table.
• Total: 270.2 kg CO₂-eq/m³ → Reduction: 18.4%.

Synergistic 70:30 (14% CD&W, 6% agri):

• Cement: 252.0 kg CO₂-eq.
• CD&W processing: 49 kg × 0.015 = 0.735 kg CO₂-eq.
• Agri processing: 21 kg × 0.025 = 0.525 kg CO₂-eq.
• Combined transport: 70 kg × 0.025 km/kg × 0.10 = 0.175 (co-located) → simplified to 0.35.
• Co-processing benefit: −2.5 kg CO₂-eq (shared equipment and logistics optimisation).
• Total: 266.8 kg CO₂-eq/m³ → Reduction: 19.4%.

Cement production dominates the footprint, accounting for 93.7–97.9% of the total impacts across all systems and confirming that clinker reduction is the primary decarbonisation lever. Waste-processing burdens are minimal—CD&W processing (1.05 kg) and agricultural processing (1.75 kg) together contribute less than 1% of the total, showing that even energy-intensive waste treatment remains environmentally favourable relative to virgin cement. Transportation is likewise negligible: with local or regional sourcing distances of 25–75 km, haulage adds under 0.2% to total impacts which justifies prioritising material production over logistics. Co-processing synergies are modest but meaningful, with a –2.5 kg CO₂-eq benefit (about a 0.9% improvement) offering a marginal additional advantage for integrated waste systems. In absolute terms, all waste-containing mixes deliver substantial reductions of 18.4–19.9%, translating to 54–64 kg CO₂-eq per m³ saved—roughly 0.5–0.6 tonnes CO₂ per 10 m³ of concrete.

Performance-Normalised Environmental Impacts (Per MPa·m³)

The critical metric for engineering materials is environmental efficiency—the environmental cost per unit of delivered performance.

Table 13. Performance-normalised carbon footprint.

System	Total Impact (kg CO2-eq/m3)	Compressive Strength (MPa)	Carbon Intensity (kg CO2-eq/MPa·m3)	Performance Efficiency vs. OPC (%)	Rank
OPC Control	331.1	55.1	6.01	0.0 (baseline)	3
CD&W System	269.5	40.8	6.61	−10.0% (worse)	7
Agricultural System	270.2	46.7	5.79	+3.7% (better)	2
Synergistic 70:30	266.8	47.4	5.63	+6.3% (better)	1
Synergistic 50:50	268.0	46.2	5.80	+3.5%	4
Synergistic 60:40	267.3	46.8	5.71	+5.0%	3
Synergistic 80:20	266.8	47.1	5.66	+5.8%	2

presents carbon intensity normalised to compressive strength, revealing a more nuanced comparative landscape.

Calculation Example (70:30 Synergistic):

• Total impact: 266.8 kg CO₂-eq/m³.
• Predicted strength: 47.4 MPa.
• Carbon intensity: 266.8/47.4 = 5.63 kg CO₂-eq/MPa·m³.
• Efficiency improvement: (6.01 − 5.63)/6.01 × 100% = +6.3%.

Performance normalisation changes the picture decisively. The CD&W-only system, despite an 18.6% absolute carbon reduction, suffers a substantial strength penalty (74.0% retention) and ends up 10.0% worse in carbon efficiency than OPC, illustrating the pitfall of judging alternatives on carbon alone without performance adjustment. By contrast, agricultural residues deliver a true environmental advantage with 84.8% strength retention and 18.4% carbon reduction, and they achieve 3.7% better performance efficiency than OPC, confirming that high-pozzolanic-activity materials are genuinely superior options. The synergistic systems are best overall, with the 70:30 blend yielding 6.3% better carbon efficiency than OPC—an improvement that may appear modest but would translate to roughly 75 million tonnes CO₂-eq saved annually if adopted at the global concrete scale (~14 billion m³/year). Notably, the rankings reverse depending on the methodology: a volume-based view (all waste systems appear roughly equal) differs sharply from the performance-based ranking (synergistic > agricultural > OPC > CD&W), underscoring the critical role of the functional unit in LCA. Finally, there are diminishing returns beyond 70:30; the 80:20 blend (5.66 kg CO₂/MPa·m³) offers only a marginal gain over 70:30 (5.63), whereas 60:40 (5.71) is clearly worse, confirming 70:30 as a robust optimum.

Sensitivity to Emission Factor Uncertainty

The performance-normalised results are subject to uncertainty in emission factors, particularly for waste processing.

Table 14. Sensitivity of performance-normalised carbon intensity to emission factor variation (70:30 synergistic system).

Parameter Variation	Carbon Intensity (kg CO2/MPa·m3)	Change vs. Baseline	Impact on Advantage vs. OPC
Baseline	5.63	0.0%	+6.3% better
Cement EF +15% (1035 kg/t)	6.44	+14.4%	+4.9% better (reduced)
Cement EF −15% (765 kg/t)	4.82	−14.4%	+7.8% better (increased)
CD&W Processing +30% (19.5 kg/t)	5.65	+0.4%	+6.0% better (minimal change)
Agri Processing +40% (35 kg/t)	5.67	+0.7%	+5.7% better (minimal change)
Grid Carbon +100% (0.90 kg/kWh)	5.69	+1.1%	+5.3% better (minimal change)
Worst-Case Combined	6.51	+15.6%	+3.1% better (still favourable)
Best-Case Combined	4.75	−15.6%	+9.6% better (enhanced)

presents sensitivity analysis varying key parameters within their uncertainty ranges (from Table 3).

Sensitivity results show that the cement emission factor overwhelmingly dominates. A ±15% change in cement EF drives a ±14.4% shift in total carbon intensity, far outweighing all other parameter uncertainties and reinforcing that the principal gains come from clinker reduction rather than improvements in waste processing. Correspondingly, uncertainty in waste processing is essentially immaterial; raising the agricultural processing EF by 40% (from 1.75 to 2.45 kg CO₂/t) alters the final carbon intensity by only 0.7%, confirming its negligible contribution. The synergistic advantage is also robust: even under worst-case assumptions (high cement EF and high waste-processing EF), the 70:30 system remains 3.1% more carbon-efficient than OPC, validating its environmental superiority across plausible parameter ranges. Looking ahead, electricity-grid decarbonisation will amplify these benefits; as grids approach ~0.10 kg CO₂/kWh by 2035 (e.g., the UK target), the already small impacts from waste processing will shrink further, widening the performance gap relative to virgin cement production, which remains constrained by fossil-fuel-intensive clinker manufacture.

Comparison with Literature Benchmarks

To validate the findings of this present study,

Table 15. Performance-normalised carbon intensity benchmarks from the literature.

Material System	Carbon Intensity (kg CO2/MPa·m3)	Source	Notes
OPC (global average)	5.8–6.2	[71]	Range reflects regional cement carbon intensity
Fly ash (30% replacement)	4.9–5.4	[72]	Mature SCM with optimised supply chains
GGBFS slag (50% replacement)	4.2–4.8	[95]	Benefits from blast furnace waste stream
Silica fume (10% replacement)	5.6–6.0	[71]	High performance but limited availability
Rice husk ash (20% replacement)	5.2–5.9	[86]	Similar to this study’s agricultural system
Recycled concrete powder (20%)	6.4–7.2	[96]	Confirms performance penalty issue
This study: 70:30 synergistic	5.63	Current analysis	Competitive with mature SCM systems

contextualises the findings within published LCA studies on cement alternatives.

From the table, the synergistic system (5.63 kg CO₂/MPa·m³) outperforms standard OPC (6.01), silica fume (5.6–6.0), and recycled concrete powder (6.4–7.2), while approaching the performance of fly ash (4.9–5.4). Although a gap remains with best-in-class slag systems (4.2–4.8), the 70:30 blend still achieves roughly 69–77% of slag’s environmental-efficiency improvement potential, with the added advantage that slag availability is constrained by steel production whereas CD&W and agricultural residues are more geographically abundant. The agricultural-only system (5.79) closely matches published rice husk ash LCA ranges (5.2–5.9), reinforcing confidence in the selected emission factors and calculation method. By contrast, the CD&W-only pathway shows a performance penalty of 6.61 kg CO₂/MPa·m³, consistent with the 6.4–7.2 range reported by Florea & Brouwers [96] which underscores fundamental limitations of recycled concrete powders without complementary chemistry. The synergistic approach effectively bridges this gap: combining CD&W with high-activity agricultural ashes removes the performance penalty while preserving carbon benefits, offering a novel contribution beyond single-waste systems.

4.3.2. Global Waste Generation Analysis

Global waste diversion potential is shown in

Table 16. Global waste diversion potential.

Waste Stream	Global Generation (Gigatonnes/Year)	Theoretical Utilisation (Gigatonnes/Year)	Realistic Diversion (Gigatonnes/Year)	Diversion Percentage (%)
CD&W	3.0	0.574	0.172	5.7
Agricultural Residues	4.2	0.246	0.074	1.8
Total Synergistic	7.2	0.820	0.246	3.4

Note: 1 gigatonne (Gt) is equal to 1 billion tons.

. From the literature review,

• Global CD&W generation: 3.0 gigatonne/year [16].
• Global agricultural residue generation: 4.2 gigatonnes/year [97].
• Construction cement consumption: 4.1 gigatonnes/year [98].

With maximum theoretical utilisation if all cement were replaced at a 20% level, then the required waste = 4.1 × 0.20 = 0.82 gigatonnes/year.

At a 70:30 ratio,

CD&W needed = 0.82 × 0.70 = 0.574 gigatonnes/year.

Agri residue needed = 0.82 × 0.30 = 0.246 gigatonnes/year.

Total = 0.82 gigatonnes/year

For a realistic implementation scenario assuming 30% market penetration due to practical constraints,

Realistic diversion = 0.82 × 0.30 = 0.246 gigatonnes/year.

CD&W diversion = 0.246 × 0.70 = 0.172 gigatonnes/year.

Percentage of global CD&W = 0.172/3.0 = 5.7%.

Agri diversion = 0.246 × 0.30 = 0.074 gigatonnes/year.

Percentage of global agri = 0.074/4.2 = 1.8%.

5. Discussion

5.1. Synergistic Combinations for Optimal Blending Ratios and Performance Enhancement

The modelling indicates that performance gains arise when the physical (packing) and chemical (pozzolanic) contributions are balanced rather than maximised independently. Within the 20% total cement-replacement window evaluated here, a 70:30 blend consistently emerges as the best compromise: it pairs a predicted packing density of ~0.641 with an 83% pozzolanic activity efficiency, i.e., near-optimal space filling without unduly sacrificing reactive silica supply. This balance reflects the complementary roles identified earlier—CD&W fines supplying nucleation/packing and agri-ashes providing amorphous SiO₂ for secondary C–S–H formalised through the modified Andreasen–Andersen packing model (with interaction coefficients in the 0.08–0.15 range) and de Larrard’s interaction corrections [51,92].

Figure 1 synthesises the mechanical implication of this balance. Relative to the OPC control (55.1 MPa; 100% retention), single-waste systems underperform: CD&W-only retains 74.0% (40.8 MPa) and agricultural residues 84.8% (46.7 MPa). Blending reverses this trend in a graded way: 50:50 delivers modest improvement, 60:40 further improves it, and 70:30 peaks at 86.0% retention (47.4 MPa); beyond the optimum, 80:20 slips slightly to 85.5%, which signals a practical upper limit to the CD&W proportion before synergy is diluted. On a rule-of-mixtures basis, the 70:30 system yields an 8.3% gain over the average of the single-waste baselines (79.4%), reinforcing that the uplift is non-additive. Mechanistically, three coupled pathways underpin the predicted uplift. First, multi-scale packing increases the paste solid fraction (PF ≈ 1.08 for 70:30) and lowers the initial capillary porosity. Second, the blend sustains a strong but not excessive pozzolanic contribution (PAI ≈ 0.807 at 70:30), which converts CH to additional C–S–H and densifies the microstructure. Third, residual interaction effects—captured by a modest synergy factor (SF ≈ 1.03)—account for non-additive behaviours such as improved dispersion/rheology that reduce the effective water demand at fixed workability. Importantly, these ingredients are embedded in a physically consistent framework in which packing sets the starting porosity state and hydration/pozzolanic reactions govern space filling over time, rather than treating packing as an arbitrary strength multiplier [51].

The optimality of 70:30 is therefore a trade-off result, not solely a peak in any single metric. Progressing from 50:50 to 80:20 marginally raises packing but erodes pozzolanic efficiency; 70:30 sits on the Pareto frontier where further CD&W addition yields diminishing returns in packing for a disproportionate loss of reactivity—hence the slight decline at 80:20 visible in Figure 1. This interpretation aligns with the packing-interaction coefficients derived from the modified Andreasen–Andersen/de Larrard approach and explains why blends with too little agri-ash under-utilise chemical densification, whereas those with too much agri-ash lose the particle-scale skeleton effects contributed by CD&W. It is also consistent with prior observations of synergy in binary agricultural-waste systems, where empirical enhancements of 5–12% were also attributed to the complementary roles of filler skeletons and reactive silica [64]. The 20% total replacement level identified as optimal corresponds with replacement thresholds reported in the literature for individual waste materials [27,99], suggesting that synergy can preserve structural-grade strength while raising overall waste utilisation.

Uncertainty analysis further tempers and operationalises these insights. For the 70:30 blend, the mean strength remains 47.4 MPa (86% retention) with a coefficient of variation ≈8.7% and a 95% interval of 39.5–55.3 MPa—comfortably above typical structural thresholds but overlapping the agricultural-only case in adverse realisations. The probability of meeting ≥40 MPa is ≈91%, and the chance of achieving ≥5% enhancement over the single-waste average is ≈74%, highlighting both its promise and the need for targeted experimental calibration. Sensitivity results point to priorities for that calibration: to refine the strength–porosity mapping (Bolomey constant) and characterise PAI for both streams, as these dominate variance; synergy-factor uncertainty is secondary. In practice, these findings motivate focusing early testing around a 70:30 ± corridor at 20% total binder replacement, with attention to (i) controlling particle size distributions to preserve the multi-scale skeleton and (ii) stabilising ash reactivity through consistent calcination/grinding protocols. Within those bounds, the modelled system offers a rational route to maintain ~85–86% of control-strength while utilising two abundant wastes, an efficiency unattainable when each stream is used alone.

5.2. Environmental Impact

Life-cycle results indicate that co-utilising CD&W with agricultural residues yields a small but real environmental edge over single-waste pathways. On a volume basis, all waste-containing mixes cut cradle-to-gate CO₂ by ~19–20% relative to OPC, with the synergistic blend having the highest reduction (20%) and a marginal incremental gain of roughly 0.7–0.9 percentage points over the better individual system—an advantage largely attributable to an integrated co-processing credit of −2.5 kg CO₂-eq·m⁻³. Cement production remains the dominant contributor (≈94–98% of total impact), so benefits are primarily driven by clinker reduction; processing and haulage burdens for the wastes are comparatively negligible at the assumed local/regional scales.

The graph in Figure 2 captures this hierarchy: both single-waste options reduce emissions substantially, yet the combined system overtakes them through process integration rather than simple additivity. Thermodynamically, the −2.5 kg CO₂-eq·m⁻³ credit reflects shared equipment use and streamlined logistics, consistent with reports that simultaneous treatment of complementary waste streams can deliver measurable energy-efficiency gains [71]. Normalising by performance sharpens the conclusion. Using a performance-based functional unit (kg CO₂-eq·MPa⁻¹·m⁻³), the 70:30 system achieves 5.63 versus 6.01 for OPC—about +6.3% better environmental efficiency—ranking synergistic > agricultural > OPC > CD&W. This confirms that modest per-m³ savings translate into more meaningful system-level advantages once strength penalties/retentions are accounted for, aligning with functional-unit recommendations for cementitious LCA [71]. Robustness checks show this advantage persists across regional cement/grid intensities (≈+5.3% to +6.5%) and remains under future grid decarbonisation scenarios, since clinker emissions dominate even as electricity becomes cleaner.

At scale, Figure 3 highlights asymmetric but favourable waste-diversion dynamics. The literature estimates place global CD&W at ~3.0 Gt·yr⁻¹ [16] and agricultural residues at ~4.2 Gt·yr⁻¹ [97], against ~4.1 Gt·yr⁻¹ cement consumption [98]. At 20% cement replacement and a 70:30 split, theoretical utilisation would require ~0.574 Gt·yr⁻¹ CD&W and ~0.246 Gt·yr⁻¹ agricultural ash; under a conservative 30% market-penetration scenario, this corresponds to diverting ~0.246 Gt·yr⁻¹ combined—about 5.7% of CD&W and 1.8% of agricultural residues. These figures suggest CD&W supply could become the limiting resource in some regions and frame a realistic order-of-magnitude for circularity benefits. Importantly, such diversion rates are within ranges recognised as meaningful for sectoral circular economy progress (3–8%) [100], and the compound effects extend beyond direct CO₂ cuts: reduced landfilling of organics, less quarrying of limestone, and lower transport burdens can cascade into broader environmental benefits [16]. The environmental signal is clear but nuanced: the synergy delivers a modest percentage gain per cubic metre, yet—because it is robust to regional contexts, grounded in clinker reduction, and supported by scalable waste-diversion arithmetic—it can accumulate to material system-level impact when deployed at scale.

5.3. Synergistic Mechanisms and Multi-Objective Optimisation

The multi-objective optimisation analysis reveals trade-off mechanisms that govern synergistic performance in CD&W–agricultural waste combinations, with the optimisation landscape demonstrating clear performance boundaries that validate theoretical predictions. As illustrated in the packing density versus pozzolanic activity efficiency plot (Figure 4), the four blend ratios exhibit distinct positioning that illuminates the fundamental tension between physical and chemical optimisation parameters. Using the internally consistent particle-packing values from Table 2 with the PAI efficiencies from Table 3, the coordinates are 50:50 (φ = 0.637; PAI = 85%), 60:40 (φ = 0.640; 84%), 70:30 (φ = 0.642; 83%), and 80:20 (φ = 0.644; 82%)—each shift toward CD&W raises φ while incrementally eroding PAI. A simple slope-of-trade-off calculation between 50:50 and 80:20 indicates +1.1% relative gain in φ against −3 percentage points in PAI, i.e., about a 2.7:1 loss in chemical efficiency per 1% packing gain. This asymmetry reflects the higher sensitivity of pozzolanic reactivity to compositional changes compared with geometric packing, consistent with cement-chemistry behaviour where reaction kinetics respond non-linearly to binder composition [101].

Within this landscape, 70:30 occupies the Pareto-efficient knee: any further increase in CD&W (e.g., 80:20) gives only marginal φ improvement (0.642 → 0.644) at a disproportionate PAI penalty (83% → 82%). Thus, no alternative blend improves one objective without degrading the other, satisfying Pareto optimality criteria [102]. The mechanistic basis for synergy is as follows.

• Size complementarity (physical): Coarser CD&W particles (~55 µm) interlock with much finer agricultural ashes (~5.5 µm), tightening the packing skeleton and lowering initial capillary porosity; this size-graded effect mirrors packing gains reported for RHA systems [99].
• Complementary reactivity (chemical): Calcium-rich CD&W phases interact with silica-rich agro-ashes to form additional C–S–H beyond the sum of individual contributions, in line with observations on recycled-powder systems [63].
• Rheology/workability balance (morphology): The angularity of CD&W is tempered by the finer ash fraction (often more equant), improving dispersion and lowering effective water demand at equal flow and thereby preserving the effective w/b in blended pastes.

These concurrent pathways rationalise the modest but real interaction uplift captured by the synergy factor, SF ≈ 1.03, i.e., ≈3% performance beyond independent packing and PAI contributions—consistent with documented non-additive effects in multi-SCM systems and with the study’s coupling of packing and hydration only through physically meaningful porosity states [90]. In sum, multi-objective optimisation confirms that the system’s “best compromise” is not the blend with the very highest φ, but the Pareto-efficient 70:30 point where incremental physical gains would cost outsized chemical efficacy. This mathematically grounded placement on the frontier aligns with the mechanistic picture above and guides mixture selection.

5.4. Implications for Sustainable Construction

The research findings present significant implications for advancing sustainable construction practices through systematic waste utilisation strategies that address multiple environmental and technical objectives simultaneously. The demonstrated ability to achieve predicted improvement and environmental benefits through synergistic combinations suggests a shift from single-waste utilisation approaches toward integrated waste management systems. This shift aligns with circular economy principles advocated by Kirchherr et al. [100] and addresses the growing demand for sustainable construction materials in developing economies where both CD&W and agricultural residues are abundant. The scalability analysis indicating potential diversion of 0.246 gigatonnes of waste annually represents a substantial contribution to global waste management objectives while reducing dependence on virgin cement production. The economic implications extend beyond direct material cost savings to include reduced waste management costs, decreased transportation impacts through co-location strategies, and potential value creation through waste stream optimisation. The regional applicability of synergistic approaches is particularly relevant for agricultural regions with significant construction activity, where both waste streams are readily available, and transportation costs can be minimised.

The technical feasibility demonstrated through mathematical modelling provides confidence for pilot-scale implementation and commercial development, supporting arguments by Wu et al. [16] for integrated waste utilisation strategies in construction applications. The environmental benefits of 20.3% carbon footprint reduction support international climate objectives while addressing local waste management challenges through productive utilisation rather than disposal. Taken together, these support SDG 9 (Industry, Innovation, and Infrastructure) via new circular supply chains, SDG 11 (Sustainable Cities) through lower-climate-impact built assets, SDG 12 (Responsible Consumption and Production) by diverting and upgrading wastes, and SDG 13 (Climate Action) by reducing clinker-related emissions.

5.5. Environmental Assessment Limitations and End-of-Life Considerations

While the cradle-to-gate LCA results demonstrate favourable carbon performance for synergistic waste systems, some limitations constrain the completeness and generalisability of these findings. Transparent acknowledgment of these limitations is essential for appropriate interpretation and future research direction.

5.5.1. System Boundary Omitted Life-Cycle Stages

A limitation of this cradle-to-gate assessment is the omission of use-phase impacts, particularly differential durability that can alter environmental comparisons over a structure’s service life [74]. Because durability governs maintenance demand, accelerated deterioration increases interventions (surface treatments, crack repair, and cathodic protection) each carrying additional burdens; indeed, Purnell [73] showed that use-phase impacts can exceed production impacts when service life differs by more than ~30%. To reflect this, the amortised carbon footprint should be expressed as $\mathrm{Amortized} \mathrm{intensity}=\mathrm{Production} \mathrm{impact}/\mathrm{Service} \mathrm{life}$ (kg CO₂-eq·m⁻³·year⁻¹), or equivalently over a common reference life ($\mathrm{L}$) as $\mathrm{Production} \mathrm{impact}\times \mathrm{L}/\mathrm{Service} \mathrm{life}$. For example, if the 70:30 synergistic system achieves only 70% of OPC service life (35 vs. 50 years), OPC’s amortised intensity is $331.1/50=6.62$ kg CO₂-eq·m⁻³·year⁻¹ (331.1 kg CO₂-eq·m⁻³ over a 50-year reference life), whereas the synergistic mix is $266.8/35=7.62$ kg CO₂-eq·m⁻³·year⁻¹ (381.1 kg CO₂-eq·m⁻³ over 50 years), yielding a ~13% worse environmental performance despite a 19.4% production-phase reduction. This illustrates the critical importance of durability validation before environmental superiority can be conclusively claimed; preliminary theoretical considerations therefore suggest mixed outcomes pending experimental evidence.

The end-of-life stage (demolition, transport, processing, and disposal/recycling) is excluded here due to profound temporal uncertainty (predicting demolition practices and recycling technologies 50–100 years ahead is speculative) alongside regulatory evolution (future waste rules, landfill policies, and circular economy frameworks) and a comparative-neutrality assumption that structurally similar concretes are demolished and processed in a like manner, causing end-of-life burdens to cancel in comparative LCA. That assumption may not hold if waste-containing concretes behave differently at end-of-life: synergistic mixes might offer a recycling advantage because their demolition waste could be more readily accepted for downcycled applications (e.g., road base), yet their heterogeneous composition (cement + CD&W + agricultural ash) might complicate higher-value recycling if, for example, ash chemistry raises contamination concerns. Consistent with Design-for-Deconstruction thinking [103], composition choices should therefore consider recyclability, and future work should test whether synergistic waste concretes enable or hinder circular end-of-life pathways.

5.5.2. Geographic and Temporal Scope Limitations

Transportation baseline LCA assumes local/regional sourcing scenarios (CD&W: 50 km, agricultural residues: 75 km) representative of densely populated regions with proximate waste sources and construction demand. These assumptions critically affect environmental conclusions (

Table 17. Sensitivity to transportation distance (70:30 synergistic system).

Transport Scenario	Total Transport (kg CO2/m3)	Total Carbon (kg CO2/m3)	Performance-Normalised (kg CO2/MPa·m3)
Baseline (local)	0.35	266.8	5.63
Regional (150 km avg)	1.05	267.5	5.64 (+0.2%)
National (400 km avg)	2.80	269.3	5.68 (+0.9%)
International (1000 km avg)	7.00	273.5	5.77 (+2.5%)

).

Calculation example (Regional scenario):

• Combined waste: 70 kg/m³.
• Average distance: 150 km.
• Transport: 70 × 0.150 × 0.10 = 1.05 kg CO₂-eq/m³.

Transport impacts remain under 3% of the total even at national distances (400 km), confirming that material production overwhelmingly dominates the footprint; the environmental advantage also persists at international sourcing scales (1000 km yields 5.77 vs. 6.01 kg CO₂-eq/MPa·m³ for OPC, i.e., +4.0% better). Conversely, optimising for short distances offers little additional benefit (only 0.2% improvement when moving from 400 km to 50 km), indicating that material quality and processing efficiency matter far more than logistics. Economically, however, transport can still be decisive: agricultural residues are bulky, low-density materials (200–300 kg/m³ in loose form) that typically require conversion to ash (≈10:1 mass reduction) to make long-distance hauling viable. This, in turn, implies the need for distributed calcination facilities near production zones and introduces regional availability constraints for practical implementation.

For regional variability in emission factors, all emission factors reflect European/UK contexts (cement: European average; electricity: UK grid; transport: Euro V trucks). Application to other regions requires adjustment (

Table 18. Regional emission factor variation and impact on results.

Region	Cement EF (kg CO2/t)	Grid EF (kg CO2/kWh)	OPC Impact (kg CO2/m3)	Synergistic Impact (kg CO2/m3)	Synergistic Advantage
EU/UK (baseline)	900	0.45	331.1	266.8	+6.3%
China	750	0.65	282.4	230.1	+5.3%
India	850	0.75	318.4	258.1	+5.7%
USA	920	0.55	339.5	273.6	+6.3%
Brazil (hydro-heavy)	880	0.15	319.7	257.0	+6.5%
Australia (coal-heavy)	900	0.85	337.1	272.8	+5.9%

).

From the table above, the synergistic advantage performance-normalised definition is

$$\mathrm{Advantage}=1-{\displaystyle \frac{I_{\mathrm{Syn}, \mathrm{region}}}{I_{\mathrm{OPC}, \mathrm{region}}\times r}},$$

where r is the 70:30 blend’s strength retention vs. OPC = 0.86.

The advantage is consistent across regions (5.3–6.5% better) using the performance-normalised metric (r = 0.86). The largest advantage appears in Brazil (+6.5%), where a very clean grid compresses the common electricity term, mildly amplifying the benefit of using less cement in the 70:30 mix. The smallest advantage is in China (+5.3%), where a high grid EF adds the same electricity burden to both systems, slightly diluting the cement-reduction benefit. Absolute footprints vary by ~19–20% between regions. The OPC range is 282–340 kg CO₂/m³ and the synergistic mix range is 230–274 kg CO₂/m³, driven by regional cement and grid intensities. Despite this spread, the ranking is stable: the 70:30 system remains environmentally preferable in every region under these assumptions.

Temporal Evolution of Grid Decarbonisation Scenarios: Electricity grid carbon intensity is declining rapidly in most developed economies due to renewable energy expansion and coal phase-out. Figure 5 projects the temporal evolution of synergistic system advantages under UK government decarbonisation scenarios. It shows declining carbon intensity for both OPC and synergistic systems with a narrowing percentage gap over time.

In the graph, absolute savings plateau at roughly 63 kg CO₂/m³ as batching and processing electricity becomes negligible relative to cement production, while the percentage advantage rises slightly—from 6.3% to 6.6%—as grid decarbonisation disproportionately benefits OPC due to its higher electricity use for grinding operations. By 2050, cement production accounts for about 98.5% of the total impact, reinforcing that clinker replacement remains the only meaningful decarbonisation pathway in deeply decarbonised electricity systems. Even under aggressive grid decarbonisation, the synergistic waste systems retain their environmental advantage which effectively future-proofs investment in waste-processing infrastructure.

5.5.3. Single Impact Category Limitation

The exclusive focus on GWP represents a limitation, as comprehensive sustainability assessment requires multi-criteria evaluation across diverse environmental impact categories [88]. Qualitative assessment of other relevant impact categories may suggest mixed implications. Impact categories where waste systems may perform better include resource depletion (abiotic depletion potential) and land use. Waste utilisation reduces limestone quarrying (17% reduction per tonne cement replaced) and aggregate extraction, preserving finite mineral resources. Also, avoiding waste landfilling reduces land occupation. For the 70:30 system diverting 70 kg waste/m³, landfill land savings = 70 kg/1200 kg/m³ (landfill density) × 10 m height = 0.00058 m² (minimal but positive). Future research should employ comprehensive LCA methodologies such as ReCiPe 2016 Midpoint/Endpoint, Environmental Footprint (EF 3.0), or TRACI 2.1 (USA). Such multi-criteria assessments may reveal trade-offs where waste systems excel in climate change but perform poorly in particulate matter or toxicity, requiring weighting and value judgments about the relative importance of different environmental impacts—an inherently subjective process requiring stakeholder engagement [104].

5.6. Technical Challenges and Barriers

Despite the promising theoretical results, several technical challenges must be addressed to enable practical implementation of synergistic CD&W–agricultural waste systems in commercial construction applications. Material consistency represents the primary technical barrier, as both CD&W and agricultural residues exhibit significant variability in composition, contamination levels, and processing quality depending on source and handling procedures. This variability challenges the consistent performance requirements of construction applications and necessitates robust quality control systems beyond those required for individual waste streams. The seasonal availability of agricultural residues compounds this challenge, creating supply chain complexities that must be managed through storage strategies or alternative sourcing arrangements. The processing technology barriers include the need for specialised equipment capable of handling diverse waste materials with different physical and chemical properties simultaneously. Current processing infrastructure is typically designed for single material streams, requiring significant adaptation or replacement to accommodate synergistic processing requirements. The co-processing benefits identified in the environmental analysis depend on successful integration of processing operations, which may require substantial capital investment and technical expertise.

Quality control challenges extend beyond individual material testing to include interaction effect monitoring and long-term performance validation, areas where current testing standards and protocols are insufficient. Standardisation barriers represent another significant challenge, as existing codes and specifications do not adequately address multi-waste systems which creates regulatory uncertainty for implementation. The economic viability challenges include higher initial processing costs for combined systems compared to individual waste utilisation, though these may be offset by improved performance and environmental benefits over the service life of concrete structures. Market acceptance barriers stem from conservative construction industry practices and limited awareness of synergistic waste utilisation potential among practitioners and specifiers. Addressing these barriers requires coordinated efforts involving research institutions, industry stakeholders, and regulatory bodies to develop appropriate standards, testing protocols, and implementation guidelines that facilitate practical application while ensuring performance and safety requirements are met.

For uncertainty-aware implementation, global uncertainty analysis shows that prediction spread is non-trivial and must inform piloting. For the 70:30 blend, the 95% confidence interval spans 39.5–55.3 MPa, so validation should anticipate a wide outcome envelope. The 26.4% probability that synergistic combinations fail to significantly outperform individual systems (enhancement < 5%) emphasises the need for experimental programs to include contingency mixtures and alternative blending strategies. The global sensitivity analysis identifying the Bolomey constant as the dominant uncertainty source (38.9% variance contribution) suggests that early-stage calibration experiments focusing on the fundamental strength-porosity relationship will provide maximum uncertainty reduction for subsequent optimisation efforts. Practical implementation of the proposed synergistic blends critically depends on rigorous durability testing (e.g., sulphate resistance, shrinkage, chloride penetration), which this theoretical study does not cover.

5.7. Model Credibility and Planned Validation

The modelling is framed to be falsifiable and calibratable. The assumptions are explicit, the verification, calibration and validation steps are specified (Section 3.3); and ablations confirm that the principal conclusions do not depend on auxiliary multipliers (Section 4.2.5). The variance-based analysis already identifies a calibration priority order—$K$ then PAIs—which will reduce predictive uncertainty most efficiently. The immediate next step is a compact validation matrix centred on the 70:30 ± corridor at 20% total replacement with the following: (i) an OPC control $w/b$ sweep for $K,n$; (ii) standardised PAI tests for the two waste streams; (iii) independent strength tests used strictly for out-of-sample validation. Success will be judged by pre-declared error and coverage metrics and by preservation of the performance-normalised environmental ranking reported in the Performance-Normalised Environmental Impacts (Per MPa·m³) Section. This closes the loop between model structure, measurable quantities, and acceptance criteria, strengthening the scientific credibility of the framework.

6. Conclusions

This study critically evaluated the synergistic utilisation of construction and demolition waste (CD&W) and agricultural residues as supplementary cementitious materials, combining a systematic literature review with mechanistic and multi-objective modelling. Across the scenarios examined, the analysis indicates that binary blends can outperform single-stream replacements when particle packing and pozzolanic reactivity are balanced rather than maximised in isolation. A modified packing framework, pozzolanic activity indices, and Pareto analysis jointly identified an optimal 70:30 CD&W-to-agricultural ratio at 20% total cement replacement. At this composition, the model predicts 86.0% compressive-strength retention relative to OPC versus a 79.4% average for the individual waste systems, an 8.3% non-additive uplift. The trade-off surface between packing density and pozzolanic efficiency shows the 70:30 blend occupying the knee of the frontier: small gains in packing beyond this point incur disproportionate losses in chemical reactivity. Mechanistically, synergy is explained by size complementarity that tightens the particle skeleton, complementary Ca–Si reactivity that forms additional C–S–H, and improved rheology at equivalent flow; together, these effects are captured by a modest synergy factor near 1.03. While the study stops short of claiming full experimental validation, it is noteworthy that the theoretical optimum identified here aligns with several independent empirical benchmarks. The 20% total replacement level sits within the overlap of experimentally reported ranges for viable agricultural residue and CD&W systems, the predicted 86.0% strength retention for the 70:30 blend remains above the 80% threshold commonly used to judge structural acceptability, and the 8.3% non-additive uplift is of the same order as the 5–12% synergy reported for other binary waste-based binders. Therefore, the present framework is quantitatively compatible with existing mechanical evidence and use it to formulate a focused validation matrix, rather than as a substitute for that testing.

Environmental findings reinforce and add nuance to the performance result. On a cradle-to-gate basis, the synergistic blend reduces embodied carbon by 20.3%, marginally exceeding CD&W-only (19.6%) and agricultural-only (19.3%) routes. When normalised by delivered strength (kg CO₂-eq·MPa⁻¹·m⁻³), the blend achieves 5.63 versus 6.01 for OPC, a 6.3% improvement in carbon efficiency; agricultural-only scores 5.79 and CD&W-only 6.61 due to strength penalties. Scaling analysis suggests a realistic order-of-magnitude opportunity: at 30% market penetration the approach could divert about 0.246 Gt of wastes annually (≈5.7% of CD&W and 1.8% of agricultural residues), though regional supply and policy will govern feasible uptake. Uncertainty was quantified to frame decision risk. Monte Carlo simulation with 10,000 iterations indicates that the 70:30 blend has a 91.2% probability of meeting a ≥40 MPa threshold and a 78.3% probability of maintaining ≥80% strength retention, with a 95% interval of 39.5–55.3 MPa. Variance-based sensitivity attributes 38.9% of output variance to the Bolomey strength–porosity parameter and 44% collectively to pozzolanic activity indices; parameter interactions contribute 19.5%. These results motivate early calibration of strength–porosity mapping and careful characterisation of waste reactivity for the specific streams and processing routes under consideration. Taken together, the study offers a transparent, physics-informed rationale for co-utilising CD&W and agricultural residues to produce lower-carbon concretes that preserve most of the structural capacity of OPC. The conclusions are deliberately pragmatic: prioritise blends near 70:30 at roughly 20% total replacement, manage feedstock quality and seasonality, and treat co-processing as a systems problem linking materials, plant configuration, and logistics. The framework also provides a reproducible pathway for screening other multi-waste combinations using performance-normalised metrics and explicit uncertainty propagation.

Three major limitations bound these conclusions. First, the analysis is theoretical: material properties were drawn from the literature and the models have not yet been verified by controlled laboratory testing or field trials. Second, the environmental assessment is restricted to cradle-to-gate boundaries and a climate-change indicator; potential trade-offs across other impact categories and end-of-life scenarios were not quantified, and regional inventories will shift absolute values. Third, despite probabilistic treatment, several inputs exert material influence on model outputs—underscoring the need for site-specific calibration before confident specification. Within these bounds, the work outlines a tractable agenda for converting predictions into evidence and for de-risking early adoption. Three areas of further research follow directly from the limitations and findings: (1) Targeted experimental validation and calibration—design a focused laboratory programme around the 70:30 ± corridor at about 20% replacement, including replicated strength testing across water-to-binder ratios, direct measurement of pozzolanic indices, determination of the Bolomey parameter for the chosen wastes, and a core durability suite covering sulphate attack, drying shrinkage, chloride ingress, and freeze–thaw. (2) Cradle-to-grave, multi-criteria life-cycle assessment—expand the LCA to include use-phase and end-of-life processes and evaluate additional impact categories using performance-normalised functional units with regionally parameterised inventories. (3) Pilot-scale co-processing and quality assurance—establish pilot production to test simultaneous handling of heterogeneous wastes, implement feedstock certification and interaction-aware QA, and collect operational data on logistics and energy use to inform codes, specifications, and procurement criteria. By evidencing synergy through modelled performance, carbon efficiency, and realistic diversion arithmetic, the study furnishes a clear basis for early pilots and policy dialogue. Adoption should proceed through performance-based specifications, transparent reporting of uncertainty, and incremental scaling that links laboratory calibration to field trials and supply-chain learning across diverse regions.