Separations and purifications are necessary for upstream processes as well as in maximising and improving product recovery in downstream processes. These processes account for a significant fraction of the total capital and operating costs and also are highly energy intensive. Consequently, a better understanding of separation and purification processes, current and possible alternative and novel advanced methods is essential for achieving the overall techno-economic feasibility and commercial success of sustainable biorefineries.
This book presents a comprehensive overview focused specifically on the present state, future challenges and opportunities for separation and purification methods and technologies in biorefineries. Topics covered include: Equilibrium Separations: Distillation, liquid-liquid extraction and supercritical fluid extraction. Affinity-Based Separations: Adsorption, ion exchange, and simulated moving bed technologies. Representation of the three-stage separation scheme for intracellular products. The process streams and tasks are shown inside the boxes representing each stage.
In stage I, the tasks of cell harvesting, cell disruption, and phase isolation are essential and must be performed in series, while pretreatment is optional and can be performed before cell harvesting or phase isolation tasks. The task in stage II is product concentration which may comprise single or multiple technologies. In stage III, purification and refinement tasks can be accomplished by either a single or combination of technology options based on the features of the input stream and final product specifications.
Technology options available for performing the tasks listed in three separation stages.
Thus, intracellular chemicals can be categorized into specific product classes based on their properties. Such classification helps in identifying the relevant tasks and technology options in each stage of the separation scheme. The potential separation stages and the relevant technology options can be reduced using additional product properties discussed earlier in intracellular product classes.
Hence, building upon the previous work on separation schemes [ 96 ] and superstructure-based synthesis of separation networks [ 97 ], we generate an appropriate separation superstructure for each class of product. The next steps are formulation of a superstructure optimization model, solution to identify the optimal separation network design, and economic assessment. The optimization model is formulated as a mixed-integer non-linear programming MINLP problem, with binary variables denoting the active 1 and inactive 0 states of technologies present in the separation superstructure.
The optimization model is formulated in GAMS After generating a separation superstructure for a product class, we have multiple technologies which can perform the same task and each technology has a performance metric which indicates its suitability over other parallel technologies.
The MINLP optimization model comprises separation technology models, stream flows, and product recovery and purity constraints. We formulate a base case using reference values for parameters such as input conditions, technology efficiencies, material cost, and requirements, and solve the optimization problem to identify the key cost drivers. However, since these reference parameters affect the process economics, we perform additional analysis to study how variations in the values of the aforementioned parameters impact the overall process cost and technology selection.
Step 1 Formulate a base case and solve it to determine the optimal separation network and the key cost contributors. Also, determine alternate next best configurations, to decide which technologies are essential and which can be changed in the optimal design with little compromise in the process cost. Step 2 Vary a combination of parameters for the key cost contributing technologies i. Step 3 Extend the analysis to other product classes based on 1 the results for the representative case, if the same technologies are available for the other classes, or 2 the literature, individual technology considerations, and simulation tools [ ], if new technologies should be considered.
The proposed analysis framework is applied to the aforementioned product class. This simplified scheme is used to generate a separation superstructure as illustrated in Fig. Since the product is intracellular, we have four tasks in stage I: 1 pretreatment, 2 cell harvesting, 3 cell disruption, and 4 phase isolation. Pretreatment is optional and can be used to increase the effective size of the cells through flocculation. Cell harvesting is used to separate the cells from water present in the bioreactor effluent stream. Cell disruption releases the desired product along with other non-product cellular materials NPCM such as proteins, carbohydrates, lipids, nucleic acids, and cell debris in the resultant stream [ , ].
Thus, the phase isolation task for this class may require multiple steps. One of the steps is product-rich phase formation that can be accomplished by two alternatives: differential digestion and solubilization. It consists of three stages distinguished using different colors : I cell and product isolation: red ; II product concentration: green ; and III product purification and refinement: blue.
- Death In Siberia;
- The Boy Recession?
- The Birth of Solidarity: The Gdańsk Negotiations, 1980.
The technologies involved are flocculation Flc , sedimentation Sdm , centrifugation Cnt,1,2,3,4 , filtration Ftt,1,2 , acid hydrolysis Ahy , enzyme lysis Ely , bead mill Bml , membrane processes [microfiltration MF,1,2,3,4 and reverse osmosis RO,1,2 ], differential digestion Ddg , solubilization Slb , precipitation Prc , and drying Dry. An option for bypassing Byp a set of parallel technologies is included in stage I.
A Review of Membrane Technology for Integrated Forest Biorefinery
Differential digestion uses an agent for NPCM digestion leaving the product unaffected in the process [ 52 , ], while solubilization [ , ] uses a solvent which can selectively dissolve the product, leaving NPCM as is in the stream. Isolation of product using differential digestion already achieves substantial product concentration, and thus stage II is not required. However, if the product has been isolated using solubilization, then stage II is required to recover the product by precipitation using an anti-solvent. This is followed by membrane separation microfiltration to separate the precipitated product from the liquid phase.
The product obtained after the two separation stages may still contain small amounts of water, acids, solvents, and anti-solvent. Drying in stage III can remove these traces and achieve the desired dry solid state with required purity specifications for the product. The base case optimal configuration and cost contributions are presented in Fig. The technologies selected in stage I include flocculation Flc for pretreatment, centrifugation Cnt,1 for cell harvesting, acid hydrolysis Ahy for cell disruption, centrifugation Cnt,2 for initial phase separation, and differential digestion Ddg for product-rich phase formation, followed by centrifugation Cnt,3 to separate the solid product from the digested NPCM and other liquid phase components.
Stage II is bypassed because of the relative product concentration already achieved in stage I. The final product refinement in stage III is achieved by drying Dry. The cost contribution shown in Fig. The relative contribution by different tasks in the separation process is also presented. The separation cost is a summation of the annualized capital, materials, consumables, utilities, labor, and other costs refer Additional file 1.
The active streams are shown by bold red lines and selected technologies are highlighted in different colors corresponding to each stage: red for stage I, green for stage II, and blue for stage III. Cost contribution shown by the numbers on the left bar indicates stage I to be the key cost driver, followed by feed cost and stage III. Stage II is absent in the optimal network. To determine the alternate next best separation configurations, we add successive integer cuts [ ] as additional constraints in our model see Additional file 1.
Based on the base case cost contribution analysis, some key process parameters are identified.
Changes in the values of these parameters have the potential of affecting the optimal separation network design as well as process economics. Biomass titer in the feed entering the separation network is a parameter dependent on the microbial strain, cultivation route, substrate utilization, and bioreactor design. It has a potential to be altered by upstream features such as metabolic engineering tools [ 23 , ]. For example, microbial strains can be engineered to enhance the accumulation of desired product inside the cells [ — ]. Additionally, suitable microbial hosts that can tolerate stress conditions such as product or co-product toxicity, and growth inhibitors can achieve high product yields and biomass titers [ ].
Thus, biomass titer can differ for different microbial strains and product systems, and hence it is selected for further analysis. It is dependent on cell size, relative cell concentration in the process stream, and properties such as hydrophobicity, relative density, and shear susceptibility [ 85 , 86 , ]. These are properties of the components in the system, and hence there is limited scope for performance enhancement when compared to centrifugation and filtration options [ 64 ]. Cell disruption technologies After cell harvesting, the subsequent step is cell disruption.
The concentration of cells in the stream entering the cell disruption technology has a considerable effect on the overall cost as well as product release. Bead mill is a mechanical method for cell disruption and does not introduce any secondary agents in the system. However, the properties like cell wall thickness, cell size, and shear susceptibility can change the product release efficiency [ 64 , , ]. Chemical and enzymatic lyses are other cell disruption methods which have gained popularity due to higher product release efficiency, selectivity, and low energy requirements.
However, these methods introduce other components in the system, which increases the amount of materials handled downstream. The potential to recover and recycle the enzymes and chemicals in these methods is an important issue which needs further research. Enzymatic lysis can result in smaller cell debris particles which can be difficult to isolate later.
Cost-efficient separation and purification
All these parameters and considerations can alter the product release efficiency of the cell disruption technologies. Hence, in the third analysis, we choose the performance of cell disruption technologies, defined in terms of the percentage release of intracellular components for the bead mill Bml , acid hydrolysis Ahy , and enzyme lysis Ely. Phase isolation technologies The performance of differential digestion and solubilization is influenced by the amounts and costs of digestion agent or solubilizing solvent added [ 52 , , ].
Thus, we select the materials added for these parallel technologies for further analysis. Consequently, we do not perform additional analysis for stage II and stage III parameters for this product class. The results from the proposed framework are presented and some insights regarding shifts in technology selection and changes in optimal separation network design are provided. We vary the biomass titer in the range of 0.
Books by Shri Ramaswamy
We observe that the optimal separation network design changes with the change in titer values. Overall process cost and contributions by feed, separation stages, and tasks with varying biomass titer. The selected technologies and active streams are highlighted in Fig. For lower biomass titers design A , enzyme lysis Ely proves to be the optimal choice as compared to acid hydrolysis Ahy in the base case design B because the enzyme required for disruption depends on the amount of biomass [ , ], while acid is added to maintain a certain normality in the aqueous phase [ ].
For titer values of 15 and greater, the microfiltration MF,2 option is selected after differential digestion Ddg as compared to centrifugation Cnt,3 in the base case design B because microfiltration MF,2 has a better retention and concentration factor as compared to centrifugation Cnt,3.
Along with the changes in optimal network design, we also observe changes in the dominant cost drivers Fig. For titers less than 2. For titers between 2. The second analysis is performed by varying the cell separation efficiency for centrifugation and cell retention factor for filtration [ 64 , 85 , ]. The results show that centrifugation is the preferred technology for biomass harvesting in most cases corresponding region shown by the vertical contour lines. Sedimentation is selected when centrifuge efficiency is less than Overall process cost with variation in performance of cell harvesting technologies.
The contour lines denote the viable region for the three technologies available for cell harvesting: centrifugation, filtration, and sedimentation. They are horizontal in the region where filtration is selected, whereas they are vertical where centrifugation is selected. The constant color rectangular region [ 70,70 , 70,82 , Critical values, when there is a change in technology selection from centrifugation to filtration or sedimentation, are shown by white lines.
Centrifugation is the preferred technology in most cases because its major cost contributor is utility, which is lower than the large capital cost of the sedimentation tank and the high consumable replacement cost of the filtration membrane. Also, in some cases, although individual centrifugation may be costly, its combination with other technologies renders a lower cost due to increased centrifugation efficiency and decreased input flow rate into other technologies.
The analysis for cell disruption technologies is performed by varying the percentage release of intracellular components for the three technology options: bead mill, acid hydrolysis, and enzyme lysis [ — ]. All the three available technologies can have variable component releases depending upon the type of microbial biomass handled, the entering feed characteristics, and product sensitivity to harsh conditions.