An overview to process design, simulation and sustainability evaluation of biodiesel production

Sustainability, as alike concept to sustainable development, has been well thought-out to encompass the primary balance of three dimensions: environmental, economic and social, where poor performance related to one could impede performance on the others [46].

Life cycle sustainability assessment (LCSA) refers to the evaluation of all environmental, social and economic impacts in decision-making processes towards more sustainable products throughout their life cycles. Initiated from life cycle assessment, the life cycle thinking approach has been extended since 2002 to form a LCSA methodology framework, which consists of three pillars (Fig. 3)—environmental life cycle assessment (LCA), life cycle costing (LCC) and social-LCA. As a systematic and rigorous evaluation framework, life cycle sustainability provides integrative and holistic perspectives for multi-criteria decision on a given process or a system. As generalized in Eq. (1), LCSA accounts for all input–output flows occurring at each life cycle stage throughout the ‘cradle-to-grave’. Formalized by the International Organization for Standardization, LCA quantifies the environmental footprints associated with all stages of a product, service or process. LCC and SLCA examine the holistic economic aspects and social consequences respectively, evaluating the improvement opportunities of various product systems and processes including biodiesel:

where the variable \(E{I}_{\mathrm{kpi}}\) denotes the total sustainability impacts of a given process expressed as key performance indicator kpi (e.g. global warming potential and economic costs). \(E{I}_{\mathrm{kpi}}\) is determined by the characterization impact factors for input resource r \({(EIf}_{r,\mathrm{kpi}}^{\mathrm{in}}\)) or emitted compound c (\({EIf}_{c,\mathrm{kpi}}^{\mathrm{out}})\) and the input or output flows \({(X}_{r,s}^{\mathrm{in}} \mathrm{or} {X}_{c,s}^{\mathrm{out}})\) at life cycle stage s.

Evaluation of sustainability aspects have increasingly been reported for biodiesel production process during the last decade. However, most of the reports focused on the environmental and economic aspects of sustainability while omitting social aspect. The following sub-sections present a detail discussion on the techno-economic and environmental performance of biodiesel production processes.

Economic performance is the most imperative factor for evaluating the sustainability of biodiesel production and plays a vital role in industrialization of any process. The higher production cost is the major challenge for biodiesel production scaling-up and its use as an alternate to petro-diesel [47]. However, an extensive research has been conducted during the past decades concerning the process economics and product cost reduction. These researches elaborated the utilization of different feedstock together with alternative technologies for the production and purification of biodiesel. Most of the studies analysed the total investment required for biodiesel production including fixed capital investment and production cost. Such cost estimation are often based on the process flowsheet and affected by the equipment type and size, construction material, material and energy balance [48]. Economic analysis can be performed in commercially available softwares such as Aspen In-Plant Cost Estimator or Aspen Icarus Process Evaluator [23, 32]. The key variables that determine the economic performance of a given biodiesel production plant include the production capacity, the type of feedstock, and the technological production process.

The production scale is the significant factor that could influence the techno-economic profiles by either decreasing or increasing the unit cost of biodiesel. This was elaborated by analysing the economic performance of biodiesel production plant with varying production capacities. One of such study was carried out by You et al. [49] for alkali-catalysed biodiesel production using refined soybean oil with three different production scales (8, 30, 100 kilo tons/year). It was concluded that higher capacity led to more attractive ARR (After-tax Rate of Return) with a lower BBP (Biodiesel Break-even Price) and higher NAP (Net Annual Profit). The author also stated that increase in plant capacity gave the same economic effects for soybean oil as well as waste cooking oil. On another hand, Apostolakou et al. [50] analysed the effect of plant capacity on the economic viability of biodiesel manufacturing using alkali-catalysed process. They found that production scale of 60 kilo tons/year is a threshold, above which, an increase in the production scale could improve the process viability since the production cost of biodiesel could be considerably reduced.

Similar result was reported by Van Kasteren and Nisworo [51] for supercritical process using used cooking oil with three different plant capacities (8, 80 and 125 kilo tons/year of biodiesel). They found that as the plant capacity increases, the biodiesel cost decreases from 0.52 to 0.17 US $/L. Glisic et al. [52] analysed the economics of the three different biodiesel production processes and investigated the effect of production scale on the net present value (NPV) of the process. The processes investigated were homogenous alkali-catalysed, non-catalytic transesterification for biodiesel production and catalytic hydrogenation process for diesel production. The authors reported that the plant capacity significantly affected the NPV of all processes. Especially for catalytic hydrogenation process, the NPV increased from 7 to 53.1 million US$ as the plant capacity were increased from 100 to 200 kilo tons/year. They concluded that plant capacity below 100 kilo tons/year (for all the investigated plants) results in negative NPV value after 10 years of project life.

Most recently, Navarro-Pineda et al. [53] assessed the economics of biodiesel production from jatropha using alkali-catalysed process. They also included the upstream process of jatropha plantation and pellet production from waste cake that is obtained from oil extraction process. The authors found that the biodiesel production cost remains constant when the production capacity was greater than 10,000 m3/year. However, at this scale, the plant expenses were greater than the plant income that can only be reversed by higher Jatropha oil yields. Similarly, Kookos [54] indicated that a plant with annual production capacity > 42,000 tons could produce economically competitive biodiesel utilizing spent coffee grounds as feedstock. As reported by Apostolakou et al. [51], the unit production cost of chemical-catalysed biodiesel decreases and can be expressed as the function of plant size. A significant decrease in production cost from 0.9 to 0.75 euros/L biodiesel was observed with the increase in production capacity from 0 to 40 kilo tons/year, which was followed by a plateau [50]. Contrarily, the total capital investment increases proportionally with production size but not linearly. Generally mass production is always cost-effective and most economical and same is the case with biodiesel. This effect of plant size on the total capital investment has been investigated in previous research [23] where similar trends were shown for co-solvent and solvent-free operation. It was observed that total capital investment varies between 10 and 60 million euros while plant size increases from zero to 1000 million kg biodiesel per year [23].

Most of the techno-economic studies concluded that the high cost of biodiesel production is mainly credited to the feedstock’s price. An economic assessment study published by Haas et al. [55] demonstrated that the biodiesel production cost increases linearly with increasing the cost of the feedstocks. They found that the cost of the feedstock is about 88% of the total biodiesel production cost. Thus, there was an increasing research attention on the low-cost feedstock as a measure to reduce biodiesel costs. However, the low-cost resource often represents low-quality feedstock, which incurs additional processing costs due to pre-treatment, separation and purification steps. For example, at industrial scale, the base-catalysed process is the most economically viable option to produce biodiesel from high-quality oils [32, 56]. However, it shifts to unfeasible solutions for low-quality oil feedstock (cheaper feedstock) containing high free fatty acids and water contents due to additional energy intensive pre-treatment requirement. A technology capable to process both low and high-quality oil feedstock without any additional pre-treatment steps offers a solution. Supercritical non-catalytic and enzymatic biodiesel production technologies are the examples of such technologies that have the ability to process low-quality feedstock without any pre-treatment requirements [19, 22, 23, 32].

The economics of biodiesel production vary with production technologies, which are driven by the number of unit operations and associated costs on equipment and energy consumption [47]. Alternatively, such economic advantages may also arise due to the relatively cheaper catalyst employed in the process. Moreover, catalyst type is highly important as it defines the type and sequence of production and purification scheme.

Table 2 compares the economic evaluation studies on different catalytic processes for biodiesel production. As previously mentioned, the alkali-catalysed process gives higher yields in shorter reaction time but it is not economically viable option when low-quality oil is considered [57]. It is limited by the saponification reaction (soup formation) that occurs between catalyst and free fatty acids, resulting in energy intensive downstream purification and making the process unprofitable. Acid-catalysed process avoids the side reactions and can esterify the FFAs to biodiesel. Zhang et al. [57] showed that acid-catalysed process could give lower production cost, lower biodiesel break-even price and better after-tax-return-rate compared to alkaline process using waste vegetable oil. However, the slow reaction rate, high alcohol requirement with larger reactor size and the corrosion problems imposed by the acid catalyst do have cost implications and makes the process economically unfeasible [22, 32, 57].

Heterogeneous acid-catalysed process could be a promising alternative with economic benefits compared to the homogenous acid-catalysed process. The techno-economic analysis performed by West et al. [22] showed that the heterogeneous acid-catalysed process has better economic performance (lower production costs and capital investment) compared to the homogenous acid-catalysed process which arises due to easy separation and recyclability of the catalyst, less corrosive nature and absence of washing steps for product purification. However, the slow reaction rate and lower biodiesel yields remain the major issues with acid-catalysed processes. These issues can be addressed by transesterifying the triglycerides with supercritical methanol. Using supercritical conditions give higher methyl ester yield in a shorter reaction time with reduced purification stages which results in very competitive biodiesel prices [32, 51] compared to previously denoted processes [22, 56]. The study carried out by Lee et al. [32] further elaborated the economic benefits of supercritical non-catalytic process by estimating the most promising values for discounted cash flow return rate (DCFRR), discounted payback period (DPP), and net present value (NPV) of the plant. However, the high alcohol requirement and extreme operating conditions (350 °C and 45 MPa) [22] makes the process energy intensive and incur considerable cost to the process.

Another perspective technology is enzyme-catalysed process that is more advantageous [23, 36] than chemical and non-catalytic processes in terms of milder reaction conditions, tolerating low-quality feedstock and easy purification of the products. The enzyme-catalysed process can also be carried out in the presence of solvent to increase the enzyme productivity. Sotoft et al. [23], demonstrated that the enzyme cost that was 50% of the raw materials cost in the absence of solvent was reduced to about 22% when t-butanol was used as a co-solvent. Although, the enzyme cost was significantly reduced but this led to the high production cost due to high energy consumption for solvent recovery. Using supercritical CO2 as a co-solvent can further improve the profitability of the process by both enhancing the enzyme productivity and eliminating the energy intensive step of solvent recovery [36]. This was confirmed by Lisboa et al. [36], reporting the production cost of biodiesel as 0.75 euro/L which is lower than the cost estimated by Sotoft et al. [23] (EUR 2.35/L of biodiesel) for solvent-free process with similar enzyme productivity and price. For low-quality oil feedstock, the enzymatic process is economically superior than the acid and alkali-catalysed processes in term of capital investment but inferior in operating cost [59]. This discrepancy was due to the high cost associated with the immobilized enzyme indicating that reusing the enzyme for several batches is needed to reduce the operating cost. Profitability of the process evaluated by net present value (NPV) for assumed interest rate of 13.5% and plant life span of 10 years showed that the enzymatic process is more economically attractive than the alkali-catalysed process [44]. Generally, the reusability of immobilized enzyme or using cheap biocatalyst (soluble or liquid lipase) are the most important aspects, improvements in which could make enzymatic process economically competitive with chemical-catalysed processes.

Life Cycle Analysis (LCA) has been widely adopted as a tool to evaluate environmental performance of any product or process. In previous LCAs (see Table 3), the inventory of biodiesel production derived from computer-aided process were fed into LCA to identify environmental hot-spots contributing to the impacts and evaluate environmental sustainability of biodiesel production. As visualized in Fig. 9, the inventory including input–output flows are associated with mid-point environmental impact categories and converted to category indictors by using defined characterization factors; the aggregated indicator results provide characterized profiles of the biodiesel systems, which can be further normalized and linked with protection areas (i.e. end-point categories including human health, ecosystem, resource depletion).

Life cycle impact assessment (LCIA) phase

Biodiesel production can be largely classified as three life cycle stages. Raw material production is the first stage, which includes cultivation, harvesting, transportation and storage of oil seed crops, as well as production and transportation of all the required chemicals. The second stage involves pre-treatment (milling, extraction and purification) of oil feedstock and conversion via esterification/transesterification to biodiesel. The third stage includes storage, distribution and transportation to petrol station, and eventual burning of biodiesel. As summarized in Table 3, LCA study conducted by Hou et al. [60] adopted a full well-to-wheel approach by including all relevant processes in the life cycle stages of biodiesel (e.g. production of chemicals and energy, feedstock cultivation and transportation, production of biodiesel and combustion of biodiesel at use phase). However, majority of the surveyed studies adopted well-to-gate approach (see Table 3) excluding the step of biodiesel distribution and end use. This approach is useful when the study is conducted to compare different technological pathways for biodiesel production, since the performance of vehicle engine does not change with the fuel combustion produced from different technological routes [61]. But, when the purpose of the assessment is to compare biodiesel with their fossil substitute, e.g. biodiesel with conventional diesel fuel, the well-to-wheel approach offers better reflection of the overall life-cycle performance where engine plays a role for exhaust gas emissions and ignition performance. Significant reductions in particulate matters, hydrocarbons and carbon monoxide emission are reported which are the profound advantages of biodiesel over conventional diesel [62].

Functional unit is another important factor which quantify the identified functions of a product system in which all the materials and energy flows and all effects resulting from these flows are related [63]. Mostly four types of functional unit can be identified in biodiesel LCA which include input-related units, output-related units, unit of agriculture land and year [64]. In biodiesel LCAs, majority of studies selected functional units based on the output of the product system (e.g. ton of biodiesel, L of biodiesel, MJ of biodiesel) [60, 65,66,67], while few studies used agricultural land and kilometres of transportation service as a functional unit [68, 69]. Besides, some studies presented the final results using multi-functional units [68, 70]. Ravindra et al. [70] used input, output and agricultural land related functional units. They used the product biodiesel as the output-related functional unit; for oil extraction functional unit is the production of 1000 kg of oil while functional unit for agriculture stage is per hectare of cropland. Similarly, Zhang et al. [69] reported two output-related functional units in their study for biodiesel based on the MJ of biodiesel and 1 km of driving distance. The implementation of kilometre of transportation service as a functional unit is better option when the goal is to compare biodiesel and fossil fuels used for transportation. Assessment with multiple functional units avoids biased outcomes and is highly effective for better assessment of any system in diverse scenarios.

Apart from functional unit and system boundary definition, the allocation approach, i.e. partitioning of environmental burdens among the multiple product is of great importance for biodiesel systems [63]. In biodiesel LCAs, the key allocation concern is between the biodiesel and by-product glycerol. There are mainly four options for adopting the allocation approach namely, null allocation, physical allocation, economic or market value allocation, and system expansion or substitution-based allocation [71]. Among the biodiesel LCAs surveyed in this review, the choice of allocation is dispersed (Table 3). The allocation adopted in most of the biodiesel LCAs were based on the physical properties of the product. Some studies related to biodiesel LCAs adopted the null-allocation approach and assigned all the environmental burdens to the main product biodiesel. However, this approach is not necessarily representative of the actual contribution of the studied products. Different allocation procedures may influence the results of biodiesel LCAs, which should be evaluated by sensitivity analyses [63]. Castanheira and Freire [72] analysed the sensitivity of the final LCA results to different allocation approaches in palm biodiesel evaluation. They adopted three different allocation methods (mass allocation, energy allocation and economic allocation) and stated that the environmental impacts estimated with energy and economic allocation were higher than those obtained with mass allocation. Our summary in Table 3 presents a lack of robustness analyses in the biodiesel LCAs, i.e. sensitivity analyses not presented in most of the published work.

A number of research articles have been published on the evaluation of environmental performance of biodiesel and its use by considering various feedstock and alternative production technologies. Following sub-sections discuss in detail the environmental performance of biodiesel utilizing various feedstock and different production technologies.

A variety of feedstock can be utilized for biodiesel production that offers environmental benefits based on their requirements for agriculture, transportation and several other conditions. The feedstock assessed for biodiesel environmental performance through its life cycle includes first, second and third generation feedstock along with waste oils and fats (see Table 3). Hou et al. [60] conducted a comprehensive LCA of biodiesel from different feedstock (soybean, jatropha, microalgae) and compared the environmental performance with conventional diesel (fossil-derived). Among different feedstock, microalgae come out as more feasible alternative in terms of terrestrial eco-toxicity potential (TEP) and fresh water aquatic ecotoxicity potential (FWAEP) due to lower agriculture inputs. Hou et al. [60] found that FWAEP that is caused by agricultural process contributed 92%, 43.9% and 91% to the total environmental burden in the life cycle of jatropha, microalgae, and soybean-based biodiesel, respectively. In comparison to conventional diesel, biodiesel performed better in terms of global warming potential (GWP), ozone layer depletion (ODP) and abiotic depletion (ADP), but showed worse performance in acidification, eutrophication, photochemical oxidation, and toxicity [60]. The better performance of biodiesel in ADP, GWP and ODP is principally due to CO2 uptake and solar energy from the environment through photosynthesis during the biomass agriculture. In another study, the environmental performance of second-generation biodiesel was compared with waste oil-based biodiesel [65]. When non-edible oil from jatropha is compared with waste cooking oil for biodiesel production, the latter showed lower environmental impact to all damage categories (climate change, human health and ecosystem quality). The inferiority of jatropha-based biodiesel in environmental performance is attributed to fertilizers, chemicals, water and land requirements for biomass cultivation and harvesting [65]. However, waste cooking oil-based biodiesel showed severe environmental impact for damage categories of resources (including mineral extraction and non-renewable energy demand). The total burden on the environment was 74% lower in case of utilizing waste vegetable oil as a feedstock compared to jatropha oil [65].

Further to compare environmental impact of a variety of waste feedstock, Dufour et al. [73] adopted well-to-gate analysis of feedstocks including beef tallow, sewage sludge, poultry fat and waste vegetable oil. The scope of the study was further extended by conducting well-to-wheel analysis of first-generation feedstock (soybean and rapeseed) to compare the impacts of waste oil derived biodiesel with first generation and conventional diesel. When these findings were compared, results elucidated the environmental superiority of FFA-rich materials derived biodiesel compared to both first-generation biodiesel and conventional diesel. While, among FFA-rich feedstock, waste vegetable oil showed better environmental performance in terms of GHG savings [73]. It can be conferred from the above discussion that waste oils are paramount encouraging feedstock for biodiesel production.

Besides comparing different potential feedstock, LCAs were also conducted on the perspective of comparing different technological pathways for biodiesel production. One of such study was conducted by Morais et al. [74] to evaluate environmental viability of biodiesel produced from three technological alternatives including non-catalytic process (supercritical) with propane as a co-solvent, acid-catalysed process, and traditional alkali-catalysed process with acid pre-treatment. For each of the alternative technology, depletion of abiotic resources and marine aquatic ecotoxicity potential were found the most relevant environmental impact categories. Methanol that is used as a raw material in all alternative processes, significantly contributed to the depletion of abiotic resources since it is synthesized from fossil resources. Compared to methanol, ethanol could be a preferred option due to its renewable origin. That is, ethanol is responsible for absorbing significant amount of CO2, decreasing significantly the GHG effect of the manufacturing system [75]. Beside this, non-catalytic (supercritical conditions) route using propane as a co-solvent is relatively more environmentally favourable process [74]. This is because of the absence of catalyst and its lower steam consumption compared to other process.

While, the acid-catalysed route generally causes the highest environmental impact, mainly due to high energy profile related with methanol recovery operation. Compared to alkali-catalysed process, the supercritical non-catalytic process was reported to reduce the acidification by 754%, abiotic resource reduction by 313%, marine aquatic ecotoxicity by 793%, and global warming by 496% [74]. When the environmental impact of alkali catalyst (potassium hydroxide and sodium hydroxide) is compared, sodium hydroxide (NaOH) exhibited greater environmental impact on ecosystem quality and human health [76]. This can be explained by the sodium hydroxide that is an environmental hazardous material as compared to potassium hydroxide (KOH). Moreover, NaOH produces water-soluble salts on neutralization with acid and KOH precipitated to potassium sulphate by reacting with sulphuric acid. Salt precipitation decrease the overall water consumption and discharge of polluted water to environment, while this is not the case in using NaOH [23].

In contrast to aforementioned studies, many researchers evaluated enzymatic technology for biodiesel production in their LCAs and reported that this technology has potentially lower environmental impact as compared to chemical catalytic technologies. For example, using biocatalyst (phospholipase) for degumming vegetable oils could reduce 44 tonnes of equivalent CO2 per 1000 tonnes of oil produced because of high efficiency and low raw material requirement [77]. To further elaborate the environmental benefits offered by enzymatic production of biodiesel, LCAs were conducted to compare enzymatic process with alkali-catalysed process. These studies showed that enzyme-catalysed process outperforms the alkali-catalysed process in each measure of potential impact categories including human toxicity, global warming, and depletion of ozone layer [33, 70]. Ravindra et al. [70] compared the results for both processes based on the single score and final total score. The single score result pointed out that, for both processes, the land use contributes the most to the environmental impact (75% for enzyme-catalysed and 70% for alkali-catalysed). However, the total score indicated less contribution to the total environmental impact by the enzyme-catalysed process [70]. Using immobilized enzyme instead of free enzyme in biodiesel production was found to further reduce the environmental burden on the processes [67]. This is because the reuse of immobilize lipase reduces consumption of minerals and carbohydrates needed for its soluble form production.

Overall, the enzymatic production technology provides significant reduction in environmental impacts compared to chemical-catalysed processes. However, photochemical ozone creation, global warming potential, terrestrial ecotoxicity and human toxicity potential are some of the impact categories in which enzymatic process shows almost same contribution as the conventional alkali-catalysed process [11]. These impact categories can be made lower for enzymatic process when the agriculture stage is avoided and a low-cost waste vegetable oil is used as a feedstock. In a study, it was estimated that for one tonne biodiesel production, 1775, 1633 and 383 kg of CO2eq is emitted to the atmosphere by alkali-catalysed, enzyme-catalysed, and enzyme-catalysed using waste cooking oil, respectively [11]. The latter process shows significant reduction in greenhouse gas emissions. Figure 10 shows greenhouse gas emissions for biodiesel in the surveyed LCA studies in this review (see Table 3). Generally, GHG emissions range from 0.51 × 10–4 to 0.11 kg CO2eq/MJ of biodiesel, which is in most cases lower than the conventional diesel ensuring net GHG reductions for using biodiesel as a substitute to petro-diesel. The variation in GHG emissions with the same technology and utilizing the same feedstock can be attributed to the variation in the system boundaries, allocation methods and other methodological assumptions. For most of the cases, enzymatic processes show considerable reduction in GHG emissions compared to chemical-catalysed processes, which is probably due to the decrease in energy consumption. Comprehensively, it is inferred that the enzymatic process is more environmental benign process as compared to the chemical-catalysed processes.

GHG emissions in surveyed biodiesel life cycle studies [11, 29, 60, 67, 73, 75] (conventional diesel [68])