Kaizen as a Circular Economy Enabler in Agro-Industrial Banana Operations: Empirical Evidence of Plastic Waste Reduction and Material Value Recovery

2.1 Circular economy for plastics: from policy aspiration to operational reality

The transition from linear to circular plastic value chains has been institutionalized as a policy priority at multiple governance levels. Foschi and Bonoli (2019), analyzing the European Strategy for Plastics in a Circular Economy, documented how policy frameworks—from single-use plastics directives to packaging and packaging waste regulations—are restructuring market incentives across the plastic value chain, from primary producers through converters, brand owners, retailers, waste collectors, and recyclers. Their analysis is particularly relevant for agro-industrial producers engaged in export markets, for whom European regulatory requirements create material compliance obligations and market access pressures toward circular plastic management.

The technical dimensions of plastic circularity have been comprehensively analyzed by Gracida-Alvarez et al. (2023) through a CE sustainability analysis framework combining life-cycle assessment (LCA) and material flow analysis (MFA). Their framework, applied to PET bottle systems, demonstrates that the choice of CE strategy—mechanical recycling, chemical recycling via enzymatic hydrolysis, or methanolysis—generates substantially different life-cycle impact profiles. For agricultural HDPE applications, the relevant implication is that mechanical recycling (the most economically accessible option for banana producers) is viable and environmentally beneficial when material contamination is controlled—precisely the operational condition that Kaizen-driven handling improvements create.

Rhein and Sträter (2021) provide critical analysis of corporate commitments to circular plastics, demonstrating that voluntary commitments by major consumer goods manufacturers predominantly emphasize recycling over the more impactful strategies of reduction and reuse, reflecting the relative operational ease of downstream recycling versus upstream material use reduction. This finding motivates the present study’s integrated approach: the Kaizen intervention pursues reduction and handling improvement (upstream CE strategies) simultaneously with segregation for recycling (downstream), targeting both the volume and quality dimensions of the waste management challenge.

Liu et al. (2018) further document the role of circular economy principles in industrial ecology frameworks that reconfigure industrial parks toward closed-loop material flows, generating both environmental and economic benefits through waste exchange and resource recovery. While their analysis focuses on industrial park contexts, the underlying logic—transforming waste from a cost liability into a resource input—directly applies to banana production operations where HDPE waste represents both an environmental burden and an untapped secondary market opportunity.

2.2 Kaizen as strategic sustainability tool: evidence base and theoretical extensions

Kaizen’s effectiveness as an operational improvement methodology is supported by an extensive empirical literature. Goyal et al. (2019), in their landmark Scopus-indexed case study of hazardous waste reduction through Kaizen in an Indian manufacturing firm, achieved 13.8% waste reduction at minimal implementation cost, establishing the foundational principle that systematic process improvement generates environmental co-benefits. Critically, their analysis documented the sustainability of improvements through cultural embedding—Kaizen’s participatory structure ensures ongoing commitment beyond initial implementation, addressing the long-term effectiveness challenge.

Sahmi and El Abbadi (2024), in a systematic literature review of Kaizen’s industrial evolution spanning 98 research articles from 2000–2022, documented Kaizen’s progressive expansion from manufacturing into diverse sectors, noting consistent waste reduction as a primary outcome alongside productivity improvement. Their analysis identified employee engagement and standardization as the most robust predictors of Kaizen effectiveness, findings directly applicable to the agricultural context where worker compliance with new handling protocols is the primary implementation variable. The systematic evidence base assembled in their review confirms robust Kaizen applicability across contexts, while also identifying agricultural applications as underrepresented—a gap the present study addresses.

Morell-Santandreu et al. (2020), in their comprehensive review of Kaizen and sustainability relationships across 172 articles, identified a critical tension: while Kaizen demonstrably generates economic and social sustainability outcomes, environmental sustainability is less frequently targeted as a primary objective, and the mechanisms linking Kaizen to environmental outcomes are undertheorized. Their findings directly motivate the present study’s explicit focus on CE outcomes as primary Kaizen implementation objectives, rather than as incidental co-benefits of efficiency improvement.

Riva et al. (2024) further document, through systematic literature review of 2017–2023 publications, that Kaizen effectiveness is amplified by digital technology integration. Their finding that digital Kaizen tools achieve up to 95% process accuracy suggests significant potential for IoT-enabled waste monitoring systems to enhance the precision and sustainability of agricultural Kaizen implementations—a research direction for future investigation.

2.3 HDPE waste in banana production: circular value recovery potential

High-density polyethylene (HDPE) constitutes the dominant plastic waste stream in banana production, generated primarily from bunch protection bags, irrigation components, packaging materials, and strapping. The material properties of HDPE—chemical resistance, mechanical durability, thermal stability, and processability—make it technically ideal for closed-loop mechanical recycling. Aguilar et al. (2021), in experimental characterization of HDPE from banana bunch covers, confirmed minimal material degradation under typical use conditions, establishing recyclability as technically feasible when contamination is controlled.

Provin et al. (2024), in their systematic integrative review of banana farming waste valorization for textile applications, document a broader landscape of circular value creation from banana agricultural residues, including fiber extraction from pseudostems, packaging material production from leaves, and polymer recycling from protective bags. Their analysis contextualizes HDPE recycling within a comprehensive banana circular economy framework, where multiple waste streams can be valorized through different circular pathways. This broader perspective reinforces the strategic value of the operational improvements documented in the present study: by preserving HDPE quality through improved handling, Kaizen implementations enable access to the most economically accessible circular pathway (mechanical recycling) while creating a foundation for more complex valorization strategies.

Chopin et al. (2016), in their sustainability assessment of banana production systems in the Caribbean, employed multi-criteria analysis to evaluate agroecological innovations across economic, environmental, and social dimensions, demonstrating that operational changes can generate significant sustainability improvements without compromising productivity. Their work provides methodological precedent for the present study’s integrated assessment approach, while also documenting the heterogeneous baseline conditions across banana farms that any replicable framework must accommodate.

The economic dimensions of banana plastic waste valorization are documented by Moreno (2020), whose analysis of a post-harvest banana bag management business model calculated that 60% of collected bags were suitable for recycling, with projected revenues of USD 105 million annually from recycled plastic sales at scale. While sector-wide projections require contextualization, the analysis documents the fundamental economic logic: HDPE waste that currently imposes disposal costs has positive secondary market value when quality is preserved—the circular value mechanism that Kaizen-driven handling improvements enable.

2.4 SDG 12 and sustainable waste management: governance framework

Sustainable Development Goal 12—ensuring sustainable consumption and production patterns—provides the overarching governance framework within which the present study is situated. Firoiu et al. (2025), in their analysis of SDG 12 achievement trajectories across EU Member States, document wide variation in progress on key indicators including waste generation intensity, recycling rates, and sustainable procurement—suggesting that even advanced economies face implementation challenges in translating SDG 12 commitments into operational outcomes. For developing-country agricultural sectors, where operational infrastructure and institutional capacity are more constrained, practical implementation frameworks are correspondingly more critical.

Ram and Bracci (2024), in their systematic literature review of waste management indicators and SDG relationships, established robust evidence for the conceptual and empirical linkages between waste management improvements and multiple SDG targets, extending beyond SDG 12 to encompass SDG 3 (health), SDG 6 (clean water), SDG 11 (sustainable cities), and SDG 15 (life on land). Their analysis underscores the multi-SDG co-benefits of agro-industrial waste management improvements, suggesting that Kaizen-driven interventions like the one documented in the present study generate sustainability value well beyond the immediate waste reduction metrics.

Rashed (2026), examining sustainable solid waste management within environmental governance frameworks, emphasizes the critical role of institutional structure and accountability mechanisms in determining waste management outcomes, documenting that technical capacity alone is insufficient absent supporting organizational and governance arrangements. This finding directly informs the present study’s emphasis on supervisory accountability structures and defined responsibilities as essential components of the Kaizen implementation framework.

3.1 Research design and epistemological positioning

This research employs an applied, quantitative approach with a pre-experimental design (O₁XO₂) and explanatory scope. The pre-experimental design enables controlled before-after comparison of waste generation metrics across the same production units, enabling attribution of observed changes to the Kaizen intervention. The quantitative approach facilitates robust statistical analysis of waste reduction across categories, subcategories, and production processes, generating empirical evidence for causal inference. The study is positioned within the positivist paradigm, prioritizing measurement precision and replicability to support generalization to comparable agro-industrial contexts.

The study was conducted between March and September 2024 in a banana export company located in Pacanguilla, La Libertad Region, northern Peru. This location is representative of Peru’s primary coastal banana production zone, supplying both domestic and European export markets, with growing sustainability certification requirements from international buyers. The company operates under Fairtrade certification requirements that create institutional incentives for documented environmental management improvement.

3.2 Population, sample, and sampling strategy

The study population comprised all active banana crops in the company during 2024. The sample consisted of 18 crops distributed across two contiguous production zones—San José de Moro (10 crops) and El Algarrobal (8 crops)—selected through purposive sampling based on representativeness of production processes and waste generation patterns. These 18 crops encompassed all six major production stages: bagging (field plastic application), de-handling (bunch separation), washing (treatment and cleaning), packaging (carton and film application), palletizing (strapping and pallet wrap), and storage/dispatch.

Sample size determination followed the criterion of operational representativeness rather than statistical probability sampling, appropriate for this organizational case study design. The 18-crop sample captures approximately 70% of the company’s total production volume and all categories of waste-generating process, enabling comprehensive waste characterization and ensuring findings reflect the company’s operational reality rather than selected production units.

Prior to data collection, written informed consent was obtained from all study participants (n = 24 workers and supervisors). The consent process included a detailed explanation of the study’s objectives, procedures, expected duration of participation, confidentiality measures, and the voluntary nature of participation, including the right to withdraw at any time without consequence. All participants signed a written informed consent form before engaging in interviews or structured observations. No data were collected from any individual prior to the receipt of signed consent.

3.3 Kaizen implementation framework: five-phase protocol

The Kaizen methodology was operationalized through five sequential, structured phases adapted to agro-industrial production conditions. Each phase incorporated specific tools, outputs, and verification checkpoints:

Phase 1 – Define: Systematic waste baseline assessment through structured field observation (three production cycles per crop), stakeholder interviews with workers and supervisors (n = 24), and initial waste audit using standardized collection and weighing protocols. Outputs: waste generation baseline by category, subcategory, and production process; identification of critical waste generation points; operational definition of waste categories for consistent subsequent measurement.

Phase 2 – Register: Development and deployment of comprehensive waste tracking systems. Outputs: standardized waste register forms at each production station; digital data collection protocols; weekly waste aggregation and category-disaggregation spreadsheets; baseline measurement period (9 weeks, one measurement per week per crop = 162 total measurement events).

Phase 3 – Design: Creation of waste classification infrastructure, color-coded container networks (7 waste category colors), process-specific handling protocol documentation, and HDPE-specific segregation protocols designed to preserve material quality for recycling. Special attention was given to the clean-contaminated HDPE separation protocol, addressing the primary quality barrier identified by Aguilar et al. (2021).

Phase 4 – Implement: Rollout of waste reduction strategies: physical infrastructure deployment (containers, signage, collection areas); supervisory accountability structure with daily waste generation review; worker training program (4-hour initial training + weekly 30-minute reinforcement sessions per production crew); process-specific minimization protocols (bag reuse optimization, stretch film usage standards, strapping recovery procedures, organic-plastic separation at de-handling).

Phase 5 – Verify: Post-implementation measurement period (9 weeks, same protocol as baseline); comparative statistical analysis; process adherence monitoring through daily supervisor checklists; HDPE material quality assessment through visual contamination scoring. Total post-test measurement events: 162, enabling paired pre-post comparison at the crop-week level.

3.4 Data collection and measurement

Waste generation data was collected at designated collection points at the end of each production cycle per crop per week. All waste was weighed using calibrated digital scales (precision: 0.1 kg; calibration verified weekly against known standards). Data were recorded across six primary waste categories: plastics (subcategorized as HDPE clean, HDPE contaminated, LDPE stretch film, PP tapes/strapping, mixed plastics), cardboard/paper, organic material, metals, hazardous materials, and other. This granular data structure enabled both aggregate effectiveness assessment and process-level diagnostic analysis.

HDPE material quality was assessed through a standardized visual contamination scoring system (1 = clean; 2 = light organic contamination; 3 = heavy contamination; 4 = co-mingled/unrecyclable), applied by the same trained observer throughout the study period to ensure consistency. This quality assessment dimension extends beyond typical waste quantity measurement to capture the circular value dimension—whether reduced waste volumes are accompanied by enhanced recyclability.

3.5 Statistical analysis

The Shapiro-Wilk test was applied to assess normality of pre- and post-intervention weekly waste generation distributions. Given confirmed non-normal distribution, the Wilcoxon signed-rank test for paired samples was employed as the primary inferential test (significance level α = 0.05). Effect size was calculated using the rank-biserial correlation coefficient (r) to assess practical significance. Descriptive statistics (mean, standard deviation, coefficient of variation) were calculated for all waste categories. All analyses were performed using IBM SPSS Statistics v.26. Multiple comparison corrections were applied for subcategory analyses using the Bonferroni procedure.

4.1. Baseline waste generation profile

Pre-intervention measurements (9 weeks) revealed total mean weekly waste generation of 587.5 kg (SD = 42.3 kg; CV = 7.2%), indicating consistent production patterns suitable for pre-post comparison. Plastics constituted the dominant waste stream at 385.2 kg weekly (65.6% of total), reflecting the material-intensive protective and packaging requirements of banana export production. Within plastics, HDPE contaminated bags represented the largest single component at 168.7 kg weekly (43.8% of plastic waste), followed by HDPE clean bags at 85.3 kg (22.1%), LDPE stretch film at 62.4 kg (16.2%), PP tapes and strapping at 45.6 kg (11.8%), and mixed plastics at 23.2 kg (6.0%). The 1:2 ratio of clean to contaminated HDPE indicated systematic quality degradation through current handling practices, suppressing recyclability and circular value recovery potential.

Process-level analysis identified bagging as the highest waste-generating operation (38% of total), followed by packaging (27%), de-handling/washing (22%), and palletizing (13%). HDPE contamination was concentrated in bagging and de-handling operations, where organic contact (soil, plant sap, residual pesticide) occurs, and where—absent systematic segregation protocols—clean and contaminated materials were co-mingled in collection. This baseline characterization established the process-specific targeting priorities for the Kaizen intervention.

4.2 Post-intervention waste reduction results

Post-implementation measurements (9 weeks) demonstrated significant and consistent waste reduction across all categories. Total mean weekly waste declined to 420.63 kg (SD = 31.8 kg), representing a 28.4% reduction equivalent to 166.87 kg/week or approximately 8,677 kg annually at this operational scale. Plastic waste declined from 385.2 to 266.88 kg weekly (30.8% reduction), with particularly strong performance in HDPE clean material (30.9%), packaging plastics (29.3%), and adhesive tapes (27.7%). Non-plastic categories showed more modest but consistent improvements: cardboard/paper −21.9%, organic waste −17.2%, metals −12.4%, hazardous materials −18.7% ( Table 1).

Beyond quantity reduction, post-intervention HDPE waste exhibited substantially improved quality characteristics. The mean contamination score improved from 2.8 to 1.9 (on the 1–4 scale), and the ratio of clean to contaminated HDPE shifted from 1:2 pre-intervention to 1:1.2 post-intervention—a transformation with direct implications for recyclability and circular value recovery. Under pre-intervention conditions, the 1:2 ratio meant that the majority of HDPE waste was unrecyclable without additional washing and decontamination; post-intervention, the majority of segregated HDPE meets standard mechanical recycling quality specifications ( Table 1).

4.3 Process-level analysis

Process-specific analysis revealed differential improvement rates reflecting the targeted nature of each Kaizen protocol. The bagging process achieved the highest reduction (32%), driven by bag reuse optimization (increasing average bag cycles from 1.8 to 2.4) and immediate post-removal segregation protocols. The packaging operation achieved 31% reduction through stretch film usage standardization and implementation of film recovery racks enabling partial reuse. De-handling and washing processes showed 28% improvement primarily through enhanced plastic-organic separation protocols. Palletizing achieved 25% reduction through strapping length optimization and introduction of strapping return bins.

The bagging and de-handling processes also showed the most pronounced HDPE quality improvements, with contamination scores declining by 1.2 and 1.1 points respectively. This reflects the effectiveness of the segregation protocols in preventing organic-plastic co-mingling at the critical points identified in the baseline assessment. The correlation between process-specific quality improvement and the baseline contamination analysis validates the diagnostic value of the Define phase in identifying high-priority intervention targets.

4.4. Statistical analysis results

The Shapiro-Wilk normality test confirmed non-normal distribution for pre-test data (W = 0.891, p = 0.023) and post-test data (W = 0.876, p = 0.014), validating non-parametric testing. The Wilcoxon signed-rank test for total waste yielded Z = -2.666, p ≤ 0.001, providing strong evidence of intervention effectiveness. Effect size calculation yielded rank-biserial correlation r = 0.89, indicating a large practical effect size—well above the r = 0.5 threshold conventionally considered ‘large.’ Across all waste categories, the null hypothesis of no difference was rejected at α = 0.05. The consistency of statistically significant reductions across all six waste categories, rather than improvement concentrated in selected categories, confirms systematic effectiveness of the intervention rather than localized effects.

5.1 The kaizen-CE nexus: Mechanism and theoretical implications

The results of this study provide empirical support for a theoretically consequential proposition: Kaizen methodology functions as a direct circular economy implementation mechanism, generating CE outcomes through three distinct but interrelated pathways that together constitute what we term the ‘Kaizen-CE nexus.’ Understanding these pathways illuminates both the theoretical mechanism and the practical conditions for replication.

The first pathway—waste volume reduction—is the most straightforward and most frequently documented in prior Kaizen literature (Goyal et al., 2019; Sahmi & El Abbadi, 2024). The 28.4% overall waste reduction achieved in this study exceeds the 13.8% documented by Goyal et al. (2019) in manufacturing, suggesting that baseline improvement opportunities are greater in agricultural settings where waste management has historically received less systematic attention—an implication supported by the contrast between the very high pre-intervention HDPE contamination rate (1:2 clean-to-contaminated ratio) and the technically achievable recycling rates documented in the literature (Aguilar et al., 2021; Shamsuddoha & Kashem, 2024).

The second pathway—material quality preservation for circular value recovery—is theoretically more significant and empirically underexplored in prior Kaizen literature. The shift in HDPE clean-to-contaminated ratio from 1:2 to 1:1.2 represents not merely an additional waste reduction metric but a qualitative transformation in the nature of the waste stream: from predominantly unrecyclable (contaminated) to predominantly recyclable (clean). This quality transformation directly addresses the primary barrier to banana plastic circularity identified by Aguilar et al. (2021) and potentially enables the closed-loop recycling value recovery documented at scale by Moreno (2020). The CE principle of ‘maintaining materials at highest value’ is operationalized precisely through this quality pathway—and Kaizen is the operational mechanism that makes it achievable.

The third pathway—systematic waste-flow data generation—is the most prospectively significant. The waste tracking systems and measurement protocols developed through the Define and Register phases create a data infrastructure that enables continuous optimization beyond the initial implementation. This data foundation also provides the quantified evidence of environmental performance improvement that sustainability certification processes and export market compliance require. Rashed (2026) emphasizes the critical role of institutional accountability mechanisms in waste management outcomes; the Kaizen framework’s supervisory accountability structures and systematic data collection effectively operationalize this governance dimension at the firm level.

5.2 Strategic positioning: from compliance cost to competitive asset

The traditional framing of environmental management as a compliance cost burden is increasingly challenged by evidence from both manufacturing (Ince et al., 2025) and service sectors (Morell-Santandreu et al., 2020) that sustainability investments generate operational efficiency co-benefits. The present study contributes directly to this reframing in the agricultural domain. The Kaizen intervention generated outcomes across multiple value dimensions simultaneously: reduced disposal costs, potential recycled material revenues, improved worker discipline and process standardization, and enhanced environmental credibility for export market certification.

For Peru’s banana sector specifically, these findings are particularly timely. Firoiu et al. (2025) document that SDG 12 targets are becoming operationalized in European import market requirements, creating material compliance pressures for agro-industrial exporters. Foschi and Bonoli (2019) further document the European regulatory trajectory toward mandated recycled content and single-use plastics restrictions that will increasingly affect agricultural packaging practices. The Kaizen framework’s capacity to generate quantified, documented evidence of waste reduction—the 166.87 kg/week reduction documented in this study—provides precisely the verifiable environmental performance data that certification processes require.

The scalability of the framework merits emphasis. Implementation infrastructure required for the Kaizen intervention—color-coded containers, signage, weighing scales, training materials—represents minimal capital investment. The high effect size achieved (r = 0.89) suggests that the improvements are driven by organizational and behavioral changes rather than capital expenditure, making the approach accessible to small and medium banana producers who constitute the sector majority but lack resources for expensive environmental management systems. Rhein and Sträter (2021) emphasize that corporate circular economy strategies are more credible when they encompass reduction and reuse alongside recycling—the present study’s integrated approach to all three, embedded in operational routines through Kaizen, provides a practical model for this comprehensive commitment.

5.3 Limitations and directions for future research

Several limitations warrant acknowledgment and motivate future research directions. The pre-experimental design, while appropriate for initial effectiveness assessment in operational settings, lacks the control groups that would strengthen causal inference. Future studies should employ quasi-experimental designs with matched control farms, enabling isolation of Kaizen effects from concurrent operational changes and seasonal variation. The 18-week study period captures immediate impacts but cannot assess long-term sustainability of improvements; longitudinal studies with 12–24 month follow-up are needed to evaluate behavioral persistence and regression risks.

The single-company, single-region sample limits external validity. Multi-site replication studies across diverse agro-industrial contexts—different crops, production scales, and governance environments—are needed to establish generalizability. Economic analysis quantifying the revenue potential of improved-quality HDPE streams would strengthen the business case for Kaizen-CE investment and enable rigorous cost-benefit analysis. Integration of life-cycle assessment (following the framework of Gracida-Alvarez et al., 2023) would enable positioning of Kaizen outcomes within broader environmental footprint and carbon accounting frameworks.

Future research should also investigate Industry 4.0 amplification of Kaizen in agricultural contexts. Riva et al. (2024) document that digital technology integration achieves up to 95% process accuracy in manufacturing Kaizen; IoT-enabled waste monitoring, digital weight capture, and AI-assisted pattern detection could substantially enhance the precision, scalability, and continuous improvement capability of agricultural Kaizen implementations—particularly relevant as precision agriculture technologies become more accessible in tropical production regions.