Enhancing Sustainability of Used Cooking Oil-Biodiesel Through Integrated Eco-Efficiency and Eco-Technology Assessment: An Indonesian Case Study

Energy security refers to the availability, accessibility, affordability, and environmental sustainability of energy use [1]. It is a critical factor in economic productivity, as disruptions in energy supply can significantly affect the production of goods and services. One key strategy to enhance energy security is the diversification of energy sources to reduce dependence on a single fuel. In this context, biodiesel has emerged as a promising alternative and remains an important component of government energy policies. Its adoption continues to expand globally, supported by increasing international recognition and policy endorsement [2].

However, the government's mandatory biodiesel policy faces significant challenges, particularly in securing supporting materials such as methanol, which is still largely imported [3]. In addition, reliance on crude palm oil as the primary feedstock creates competition with its use in the food sector, requiring careful management. The increasing biodiesel blending mandate is expected to further raise demand for crude palm oil, raising concerns over deforestation, greenhouse gas emissions, biodiversity loss, flooding, and broader socio-ecological impacts that contradict the objective of environmentally sustainable biofuels [4]. Despite these efforts, biodiesel production targets have not yet been achieved, as shown in Figure 1, indicating limitations in current production capacity. Therefore, the biodiesel industry requires alternative feedstock strategies to support the long-term sustainability of the mandatory biodiesel program.

tu_pian_1.png

Figure 1. Target and realization of Indonesian biodiesel production

In light of these challenges, exploring alternative feedstocks beyond palm oil is essential. Used cooking oil (UCO) has gained increasing attention as a viable feedstock for biodiesel production and, more recently, sustainable aviation fuel, supporting global efforts to reduce carbon emissions in the transportation sector. Global projections indicate a steady growth in UCO supply, with Asia as the main supplier and Europe as the largest consumer of low-carbon fuels [5]. In Indonesia, UCO is primarily generated from the food sector, including restaurants, hotels, and street vendors; however, its utilization requires pre-treatment to reduce water content and free fatty acids to meet quality standards [6]. UCO is classified as an advanced feedstock because it does not compete with food resources while helping to mitigate environmental impacts from improper waste oil disposal [7]. Therefore, UCO offers significant potential to reduce greenhouse gas emissions, support the circular economy, and enhance energy independence, particularly in developing countries with high cooking oil consumption [5, 8].

According to the International Council on Clean Transportation (ICCT), Indonesia has the potential to produce 157 million liters of UCO from restaurants, hotels, and schools in urban areas, while household sources could reach 1,638 million liters and are expected to increase with rising cooking oil consumption [9]. This indicates that Indonesia has strong potential for the development of UCO-based biodiesel, supported by high consumption and a widespread culinary sector. Despite this potential, a significant portion of UCO is still exported rather than utilized domestically. In contrast, countries such as India have implemented policies to promote UCO utilization through pricing mechanisms and incentives [10]. Therefore, further research is needed to assess the readiness of large-scale UCO utilization for biodiesel production, particularly in terms of process, technology, and environmental performance.

The concept of eco-efficiency serves as an important analytical approach that integrates the goals of reducing environmental impacts while improving economic performance [11]. In the context of UCO-based biodiesel, previous studies have identified electricity and methanol consumption as the main contributors to environmental impacts, highlighting the importance of optimizing these inputs to enhance eco-efficiency [12]. Furthermore, the application of a cradle-to-customer approach has been recommended to evaluate energy efficiency and emissions, supported by sensitivity analysis of factors such as transport distance and raw material variability [13]. In this regard, life cycle assessment (LCA) has been widely applied to evaluate the environmental performance of biodiesel production from UCO, demonstrating its potential to mitigate climate change, although current systems are not yet fully sustainable [14, 15].

Eco-technology refers to the application and development of technologies aimed at conserving resources, reducing emissions, and recovering material value within process systems [16]. It encompasses innovative services, processes, and systems that support economic growth while minimizing environmental impacts [17]. Eco-technology also promotes the use of non-fossil resources to address pollution challenges [18]. Compared to conventional approaches, eco-technology offers advantages in reducing environmental burdens, although it is often associated with higher implementation costs [19]. Therefore, integrating eco-technology with eco-efficiency is essential to achieve both economic and environmental sustainability. While eco-technology has been widely applied in sectors such as water treatment [20], biogas production [21], solar cells [22], and agriculture [23], yet its application in UCO-based biodiesel remains significantly constrained.

In the transition toward a low-carbon economy, the integration of eco-efficiency principles with eco-technology plays a crucial role in reducing resource intensity while improving economic performance. Accordingly, recent research has focused on developing technological frameworks that not only meet biodiesel quality standards but also optimize the ratio of outputs to resource and energy inputs. Previous studies indicate that production profitability is strongly influenced by raw material prices and operational efficiency. Techno-economic assessments further show that production costs are sensitive to pre-treatment processes, as well as methanol and catalyst consumption, while also highlighting opportunities for improving energy utilization [24]. In addition, process intensification, waste recycling, and energy recovery have been identified as effective strategies to enhance the value of UCO-based biodiesel production [25]. Therefore, improving process efficiency through better pretreatment, robust catalysts, and recovery systems remains a key opportunity for sustainable biodiesel development.

This study integrates eco-efficiency and eco-technology within a unified evaluation framework, employing LCA to assess environmental impacts and data envelopment analysis (DEA) to quantify relative efficiency among decision-making units (DMUs). Although its application in UCO-based biodiesel production remains limited, operational data such as energy consumption, chemical use, and emissions reported in previous studies can be used to construct DEA variables for evaluating eco-efficiency [12, 13]. The DEA approach offers several advantages, including the ability to compare efficiency across units, incorporate both environmental and operational indicators, identify improvement strategies, and detect inefficiencies such as wasted energy [26]. This integrated approach is consistent with recent research trends that emphasize the simultaneous evaluation of environmental and economic performance in the biofuel sector [27]. Therefore, this study aims to optimize biodiesel production by integrating eco-efficiency and eco-technology to minimize environmental impacts while improving operational performance. In addition, the study evaluates practical technology options under real operational conditions to support implementation in small- to medium-scale UCO-based biodiesel production systems.

3.1 Life cycle analysis

3.1.1 Goal and scope

The first step in conducting a LCA is determining the objectives and scope of the study, a practice foundational to ensuring completeness and accuracy, as reinforced in recent LCA guidelines for biofuel systems [30]. Therefore, the objectives and scope must be clearly defined: in this study, the objective is to identify the potential environmental impacts of the biodiesel production process using UCO from raw material arrival to final biodiesel distribution a cradle-to-customer approach similarly applied in UCO-based LCA research. This study aims to determine each process stage’s contribution to total environmental impact, with the scope focusing on energy use, raw materials, and emissions at each process stage, consistent with comprehensive impact assessments in contemporary LCA studies of waste feedstock biodiesel. The analysis includes cradle-to-customer stages raw material collection and transport of UCO, filtration, heating, mixing, settling, methanol removal, dry washing, and final delivery using a functional unit of 950 liters of biodiesel, a standard for process comparability in biodiesel LCA. The level of detail quantifying energy consumption, chemical usage (methanol and catalyst), transport distance, and waste/emissions is supported by the integration of field-collected primary data and secondary sources such as the Ecoinvent database to fill data gaps, in keeping with best practices in recent renewable fuel LCA methodologies [30].

3.1.2 Inventory analysis

Inventory analysis is conducted based on material inputs and outputs within the system. Input data consists of raw material requirements, energy/electricity, water, and transportation equipment used a structure consistent with recent cradle-to-customer LCI studies in biodiesel systems. Inventory of biodiesel production from UCO is conducted based on these inputs and outputs within the system using a cradle-to-customer approach and a functional unit of 950 liters of biodiesel produced and ready for shipment. Input data includes UCO, chemicals (methanol and catalyst), electricity and/or fuel for heating and equipment operation, and transportation during the collection and distribution stages such input categories mirror those detailed in systematic LCA inventories [31].

Inventory output data include the main biodiesel product and by-products such as crude glycerol—whose generation rate and composition (typically ~10% w/w and containing residual methanol, catalyst salts, and organic impurities) are well documented and require downstream purification or valorization strategies [32]. Solid residues from filtration and liquid wastes from washing/cleaning (including spent catalyst solutions) are common process wastes in UCO transesterification and have been quantified and discussed in recent LCA and process-optimization studies as important waste streams that affect overall environmental performance [30]. Air emissions from fuel combustion for transport and heating—primarily CO₂, with notable contributions to NOₓ and SO₂ depending on fuel type and combustion conditions—are repeatedly identified in biodiesel LCA literature as key impact drivers and should be included in the inventory with appropriate emission factors [33]. Table 1 therefore presents the complete inventory for producing of biodiesel from UCO, compiling these product, by-product, waste, and emission flows used as input for the LCIA.

To improve the traceability of material flows and ensure consistency with the defined functional unit, a mass balance was established for the biodiesel production process. It is displayed in Table 2. The balance summarizes the conversion of feedstock into biodiesel and by-products while accounting for process losses and auxiliary materials. This presentation facilitates the interpretation of the inventory data and demonstrates the consistency of material flows throughout the system boundary.

3.1.3 Impact assessment

In this study, the impact assessment was conducted using the EDIP 2003 method. The Life Cycle Impact Assessment (LCIA) stage aims to assess the potential environmental impacts arising from resource use and emissions identified in the LCI stage [34]. This process includes an evaluation of ecological impacts, impacts on human health, and the implications of natural resource utilization, with recent methodological advances (e.g., updated toxicity characterization and inclusion of near-field/far-field exposure pathways) improving how human toxicity and ecotoxicity are quantified in LCIA [35] and with dynamic/temporal aspects of LCI–LCIA increasingly recognized for capturing variability in key drivers such as electricity mixes and emissions over time [36]. The LCI therefore serves as the necessary basis for linking a product or process to its potential environmental impacts, while LCIA method choice, characterization factors, and treatment of uncertainty critically determine how those inventory flows are converted into meaningful impact results (recommendations and updated CFs / normalization guidance discussed in recent LCIA technical reviews). LCIA generates a network to represent the flow of processes, materials, and energy throughout a product’s life cycle, a visualization style commonly used in recent LCA studies for spotting hotspots and dominant subprocesses [37].

Table 1. Life cycle inventory of used cooking oil (UCO)-based biodiesel production

Table 2. Mass balance of used cooking oil (UCO) biodiesel production

3.1.4 Normalization

The normalization stage helps place LCA results into a clear context, specifically showing how significant the impact is relative to total annual emissions or environmental burdens typical for a region or population. Normalization, which converts the characterization results into Person Equivalent (PE) or milli-PE (mPE) units using the EDIP 2003 normalization factor. Figure 3 displays the normalized results for the entire biodiesel production process, illustrating how normalization can highlight critical environmental hotspots like transportation; similar findings were reported in LCA of regional biofuel supply chains where transport contributed disproportionately to global warming impacts [38]. In addition, Foteinis et al. [12] asserted that the transportation of raw materials significantly influences LCA results in UCO-based biodiesel production. These results demonstrate that the largest impact originates from transporting raw materials from the Kuta area—an observation aligned with studies showing that in biodiesel production, the transport stage often dominates the global warming potential when distribution distances are substantial [34].

tu_pian_4.png

Figure 3. Normalization life cycle assessment (LCA) of used cooking oil (UCO)-based biodiesel production

3.1.5 Single score

The following Table 3 displays the environmental impact analysis results of UCO biodiesel production, where all impact categories are converted into single-point units (mPt) to summarize the total environmental burden. It is similar to practices employed in recent LCA studies emphasizing endpoint aggregation for easier interpretation [30]. The single score generated from the LCIA analysis was then used as input for the DEA to evaluate the efficiency of the production system [39-41]. The results of the LCA show that UCO-based biodiesel provides superior environmental performance compared to first and third biofuel generation [12].

Table 3. Environmental impact value of used cooking oil (UCO)-based biodiesel production

3.2 Data envelopment analysis results

The first step in DEA is to determine the inputs and outputs for each DMU. Therefore, the specification of inputs and outputs must be clearly defined. This study aims to measure the efficiency of resource utilization in the conversion of UCO into biodiesel, identify inefficiencies, and provide improvement recommendations to optimize the production process. Furthermore, the study compares the efficiency level of biodiesel production with a scenario in which the technology can achieve higher efficiency.

Table 4. Proposed objective, input, output and decision-making unit (DMU) of used cooking oil (UCO)-based biodiesel production

Based on Table 4, the study employs seven input and one output variables. The environmental impact, which was previously considered an output in the LCA analysis, is treated as one of the inputs in the DEA, reflecting the growing use of combined LCA- DEA frameworks to capture both environmental burdens and production efficiency in a single evaluation system [39]. Other input variables include the price of UCO, the price of methanol, the price of KOH, the consumption of methanol, the consumption of KOH, and energy. Meanwhile, the volume of biodiesel produced is defined as the output. This study covers 18 DMUs, representing 18 different data collection periods.

Subsequently, the study will evaluate and rank the efficiency of processing UCO into biodiesel, as illustrated in Figure 4. The results of the efficiency score analysis showed a variation in relative performance between DMUs in utilizing resources to produce biodiesel output. A total of three DMUs (DMU15, DMU16, and DMU18) managed to achieve a score of 100%, which puts them in the frontier efficient category. These units can be considered as benchmarks and peer references because they are able to operate at the frontier of efficiency. Most DMUs achieved efficiency scores ranging from 91% to 99%, indicating that they operated close to the efficient frontier but had not yet reached full efficiency. Only three DMUs (DMU15, DMU16, and DMU18) achieved an efficiency score of 1.00 and were therefore classified as frontier-efficient under the CCR DEA model. This shows that although it has not fully reached the frontier, the majority of DMUs have been operating quite close to optimal conditions. However, there are 3 DMUs that are still in the moderate category, such as DMU4 (94%), DMU6 (87%), and DMU12 (92%). This indicates that there is a significant potential for a significant reduction in inputs, especially in energy and chemicals. This score distribution confirms that the relative efficiency between DMUs is not uniform, but is generally dominated by units that are close to the frontier. Thus, DMUs that are classified as moderate need to be directed to imitate the input usage pattern of frontier efficient DMUs. Through this benchmarking approach, it is hoped that an increase in collective performance can be achieved as well as a reduction in costs and environmental impact of the entire biodiesel production system.

tu_pian_5.png

Figure 4. Efficiency scores of the used cooking oil (UCO) to biodiesel production process

In Table 5, the results of the classification of the causes of inefficiency show that the majority of DMUs face problems with the excessive use of methanol, energy, and catalysts. In detail, as many as 5 DMUs (27.8%) experienced inefficiency due to excessive methanol consumption, such as DMU1–DMU4, and DMU12. Meanwhile, 8 DMUs (44.4%) faced problems with excessive energy use, as seen in DMU5, DMU6, DMU7–DMU9, DMU11–DMU14. In addition, 10 DMUs (55.6%) were shown to have excess use of KOH catalysts, which were spread across almost all inefficient units, including DMU1–DMU4, DMU6–DMU14, and DMU17. The DMU15, DMU16, and DMU18, which each achieved a score of 100%, did not show any cause of inefficiency, so it could be used as a benchmark. This finding confirms that the dominant factor causing inefficiency is excessive consumption of catalysts, followed by energy use, and then methanol. Therefore, improvements need to be made to achieve maximum company efficiency, one of which is by simulating improvements in the factors that cause inefficiency in each DMU.

Table 5. Classification of causes of decision-making unit (DMU) inefficiency

The authors would like to express their deepest gratitude to Lengis Hijau and PT Hijau Daun Energi for providing them with the chance, support, and access to incalculably significant data and information during the study process. Furthermore, the authors would like to thank the Ministry of Education, Science, and Technology for providing financial assistance and support under the 2025 Regular Fundamental Research Scheme.