The potential of organic solvent nanofiltration processes for oleochemical industry
Introduction
Challenges for the chemical industry increase steadily due to depletion of fossil resources, climate change and environmental regulations. However, recent trends in science and technology are also driven by an ever-increasing environmental awareness. Therefore, renewable raw materials are promoted to enhance the ecology of industrial processes. Vegetable oils and fats rank among the most important renewable feedstocks, allowing for the production of a wide variety of chemical building blocks, including fatty acids, fatty acid esters, fatty alcohols and fatty amines, which can be further processed to e.g. detergents or coatings [1]. Furthermore, fatty acid esters are directly used as biodiesel. To improve the economics of oleochemical processes, especially biodiesel production processes, the usage of non-refined oils of lower quality such as crude non-edible oils or waste oils (e.g. waste cooking oils, animal fats) is attracting attention. The application of waste oils enables raw material cost savings of up to 65% compared to refined vegetable oils [2]. Furthermore, the conversion of waste materials to fuels and valuable chemicals significantly improves the environmental impact of chemical production processes. However, the main drawback of these low-quality oils is the high content of free fatty acids, which may vary from 15 wt% for crude non-edible oils [3] to 34 wt% for waste cooking oil [4] and even up to 50–60 wt% for animal fats and tall oil [5]. This high content of free fatty acids prevents a direct application of these oils in food and chemical production processes. A further challenge to process these feedstocks is their fluctuating composition that necessitates highly flexible and efficient pretreatment processes.
Organic solvent nanofiltration (OSN), which is a pressure-driven membrane process based on solvent resistant membranes, is a promising technology for the development of energy-efficient and sustainable processes. OSN specifically targets the separation of molecules with a molecular weight between 200 and 1000 g/mol, matching the molecular weights of free fatty acids and triglycerides with chain length between 12 and 20 carbon atoms. The absence of phase transition and the mild operating temperatures of the pressure-driven membrane process result in low energy requirements compared to conventional separation technologies like distillation. Furthermore, the modular design of membrane processes provides the potential for a very flexible process design. The separation capacity, determined by the installed membrane area, can easily be adapted to the process requirements and feed conditions by modifying the number of installed membrane modules. Moreover, solid waste generation may be reduced and higher productivity-to-volume ratios can be reached at larger scale [6].
The application of OSN has already been investigated for various fields in the refining of edible vegetable oils, for example, degumming, decolorization, deodorization, deacidification and the recovery of extraction solvent [7], [8]. However, most of the available work has been focused on the processing of refined vegetable oils, while little attention was paid to the processing of low-quality oils and the challenges mentioned before. This study, therefore, focuses on two applications of OSN for the processing of low-quality oils, namely solvent recovery and deacidification. Both are essential in the production of oleochemical products. In general, edible and non-edible vegetable oils are obtained by solvent extraction from oilseeds, in which energy-intensive technologies as evaporation and vacuum distillation are used to recover the high amounts of extraction solvent [6]. Subsequent to solvent recovery, the crude vegetable oils are further refined by precipitation and centrifugation or vacuum stream stripping to achieve deacidification [9]. These processes are energy intensive, high in waste production and result in considerable oil losses [10]. By applying OSN the energy consumption and waste production in solvent extraction can considerably be reduced [11]. Furthermore, OSN might be integrated in existing processes to increase the capacity. Besides its energy efficiency, OSN is especially attractive for oil refinement because of the moderate operating temperatures that minimize thermal damage of the oils. OSN cannot only be applied for the refinement of non-edible oils for e.g. food applications, but is also a very promising technology to pretreat low-quality oils for chemical production processes. A particularly interesting application is the processing of waste oils, which attracted much attention in recent years as feedstock for the production of biodiesel. Significant cost savings are expected from a membrane-based separation compared to conventional chemical treatment by neutralization or esterification [12].
The feasibility of solvent recovery and deacidification of specific vegetable oils by OSN has already been investigated in lab scale experiments [13], [14], [15]. Our own experimental investigations have successfully demonstrated the potential of OSN for the processing of a wide range of low-quality oils [16]. High n-hexane fluxes combined with high rejection of triglycerides and fatty acids were achieved in solvent recovery, while very high rejection of triglycerides but low rejection of free fatty acids were achieved for deacidification using different polymeric membranes. Despite these promising experimental results at lab-scale, only process design calculations and large-scale pilot testing allow for a genuine comparison with established process variants. Such studies are however rare. Köseoglu et al. [17] investigated solvent recovery from oil miscellas using hollow fiber membranes demonstrating the feasibility of efficient ethanol recovery at pilot scale. However, no suitable membrane for the separation of n-hexane, which is the most common solvent in oil extraction [18], was reported, since either permeate flux or oil retention was rather low. Tres et al. [19] presented moderate to high oil rejection for the separation of soybean oil from n-butane using hollow fiber ultrafiltration membrane module, while no degradation of the membrane was found during operation. First steps towards process design were done by Raman et al. [20], who presented potential process flowsheets. For solvent recovery and partial deacidification a two stage membrane process was proposed, in which the amount of n-hexane that needs to be evaporated might be reduced by 55%, thus, leading to energy savings or increased capacity. For deacidification a multi-stage membrane cascade operated in diafiltration mode was suggested. Neither detailed process design and quantitative evaluation of energy demand nor production cost were performed. Since this is, however, crucial for judging on the potential application of OSN in the considered production processes, the current study addresses this issue in detail.
In order to determine if OSN is competitive to conventional processes, we present a comprehensive computer-aided process design and analysis. Different configurations of membrane cascades are considered as well as the combination of OSN with other unit operations in form of hybrid processes. The performance of the OSN processes is evaluated using a detailed membrane cascade model, which accounts for hydrodynamic conditions in industrial membrane modules and driving force reducing effects. A suitable local mass transfer model, capable of accurately describing the membrane performance, is incorporated and were developed in preceding work [16]. Justified evaluation of the OSN process and comparison to the reference process in terms of energy requirements, product purity and costs, should be based at an optimized membrane cascade design. To determine an optimized design a multi objective optimization approach is used. After introducing the mathematical model and the optimization approach in Section 2, the solvent recovery in oil extraction processes is described in Section 3, while the deacidification of low-quality oils for efficient biodiesel production is described in Section 4.
Section snippets
Modeling and optimization of organic solvent nanofiltration processes
In order to analyze and optimize the different process options a process model has been implemented in Aspen Custom Modeler® (ACM), allowing for the direct computation of thermodynamic and physical properties by means of available Aspen Properties® routines. The description of membrane separation performance is based on a local mass-transfer model (cf. Section 2.1). The OSN process model is augmented with additional equations accounting for transport limitations and hydrodynamics in the
Case Study 1: Solvent recovery
In the extraction process of non-edible vegetable oils the utilized high solvent amounts have to be recovered subsequently. In conventional solvent recovery processes evaporation and distillation are applied, consuming large amounts of steam and cooling water. As our previous experimental investigations have shown, OSN is a promising alternative for energy-efficient recovery of extraction solvents, being applicable to a wide spectrum of non-edible oils with varying initial free fatty acid
Case Study 2: Deacidification
Our experimental investigations have shown that OSN is a promising technology for the deacidification of non-edible and waste oils, showing high flexibility to process different oil feedstock with fluctuating compositions [16]. Therefore, OSN could be applied to purify non-edible oils, e.g. for the food industry, or used as pretreatment step in oleochemical processes to refine low-quality oils for the production of valuable chemicals and fuels. One possible application would be the upgrading of
Conclusion
In this study the potential of OSN to intensify industrial oleochemical processes of solvent recovery as well as deacidification of low-quality oils has been investigated by means of multi-objective optimization studies building on previously determined local mass transfer models derived from experimental investigations [16]. On the basis of an extended process model for a multi-stage membrane cascade design, an optimization-based process analysis reveals that for solvent recovery energy
References (58)
- et al.
Biodiesel production from waste cooking oils
Fuel
(2008) - et al.
Biodiesel production from crude Jatropha curcas L. seed oil with a high content of free fatty acids
Biores. Technol.
(2008) - et al.
Producing biodiesel from high free fatty acids waste cooking oil assisted by radio frequency heating
Fuel
(2010) - et al.
New approaches in deacidification of edible oils – a review
J. Food Eng.
(2005) - et al.
State of art of the application of membrane technology to vegetable oils: a review
Food Res. Int.
(2009) - et al.
A review analyzing the industrial biodiesel production practice starting from vegetable oil refining
Appl. Energy
(2012) - et al.
Insight into the transport of hexane–solute systems through tailor-made composite membranes
J. Membr. Sci.
(2004) - et al.
Deacidification and solvent recovery of soybean oil by nanofiltration membranes
J. Membr. Sci.
(2013) - et al.
Statistical-mechanical theory of membrane transport
J. Membr. Sci.
(1990) - et al.
A transport model for organophilic nanofiltration
J. Membr. Sci.
(2006)