Optimal Design and Synthesis of Sustainable Integrated Biorefinery for Pharmaceutical Products from Palm-Based Biomass

  • Shi Yee Ng
  • Sze Ying Ong
  • Yee Yin Ng
  • Alexander H. B. Liew
  • Denny K. S. Ng
  • Nishanth G. Chemmangattuvalappil
Original Research Paper


In the last decade, numerous technologies have been developed to convert biomass into value-added products (bioenergy, biomaterial and biochemical). However, not much research has been done in the identification of possible pathways to convert biomass into pharmaceutical products. This research focuses on exploiting the potential pharmaceutical products that can be derived from palm-based biomass. However, due to the large number of potential products, multiple reaction pathways and processing technologies involved. Thus, there is a need for a systematic methodology which is capable to identify the optimal production routes in the integrated biorefinery based on different optimisation objectives. In this work, a mathematical optimisation-based approach is developed to determine the optimum conversion pathway that converts palm-based biomass into pharmaceutical products with maximum economic performance. Besides, a novel approach which can estimate the operating cost of pharmaceutical products is also introduced in this work. In addition, sensitivity analysis is carried out to investigate the impact of changes in conversion of reaction, market price and operating cost to the economic performance of the synthesised integrated biorefinery.


Integrated biorefinery Pharmaceutical products Palm-based biomass Reaction pathway Economic performance Uncertainty 



Conversion pathway from biomass to intermediate 1


Conversion pathway from intermediate 1 to intermediate 2


Conversion pathway from intermediate 2 to intermediate 3


Conversion pathway from intermediate 3 to final product


Flow rate

\( {B}_b^{\mathrm{Bio}} \)

Total flow rate of biomass feedstock

\( {T}_n^{Inter} \)

Total flow rate of intermediates

\( {T}_p^{Prod} \)

Total flow rate of final products


Conversion rate for a given pathway


Gross profit of the overall integrated biorefinery configuration


Total annualised cost


Total annualised operating cost

\( {G}_p^{\mathrm{Prod}} \)

Cost of final products p

\( {G}_b^{\mathrm{Opr}} \)

Operating cost for conversion of biomass b

\( {G}_{s1}^{\mathrm{Opr}} \)

Operating cost for conversion of intermediate 1 s 1

\( {G}_{s2}^{\mathrm{Opr}} \)

Operating cost for conversion of intermediate 2 s 2

\( {G}_{s3}^{\mathrm{Opr}} \)

Operating cost for conversion of intermediate 3 s 3


Lignocellulosic biomass is an organic material which mainly consists of cellulose, hemicellulose and lignin. Such biomass exists in different forms such as energy crops, industrial, agricultural and forestry waste, municipal solid waste, etc. Without a proper management of such biomass, it may cause several environmental issues such as air pollution, climate change, water and soil contamination. In order to reduce the environmental impact, one of the most promising approaches in managing such biomass is to convert them into value-added products, such as biochemicals and bioenergy.

Biomass has been identified as one of the promising solutions in production of various value-added products such as biomaterials, specialty chemicals, pharmaceuticals and energy. Some of the key applications include feedstock for heat and power generation (Mohan and El-Halwagi 2006), biodiesel production (Pokoo-Aikins et al. 2009), bioalcohol production (Gnansounou and Dauriat 2010) and methanol production (Ravaghi-Ardebili and Manenti 2015). Besides, many researchers have also identified biomass as one of the prospective alternative in meeting chemical and energy requirements, improve sustainability and minimise the environmental impact which is caused by the intensive consumption of fossil fuel (Naik et al. 2010).

Since biomass is a renewable and environmental friendly resource, the establishment of an integrated biorefinery plays a significant role to ensure a sustainable supply in energy, biochemicals and biofuels (Fernando et al. 2006). Integrated biorefinery is an integrated system with various technologies which converts biomass into fuels, power and value-added products through physical, biological/biochemical and thermochemical conversion processes (Huber and Corma 2007). Based on the previous works, many biomass conversion technologies are getting mature and have been commercialised in industry such as pyrolysis, gasification, combined heat and power facilities as well as anaerobic digestion (Kimble et al. 2008). However, the abovementioned technologies are focused in the production of bulk chemicals. It is noted that such biomass can be converted into high value products such as pharmaceutical chemicals. However, there are no commercially viable pathways of a pharmaceutical product being produced directly from biomass. Therefore, biomass can be first converted into intermediates and later further converted into final product. For example, Du et al. (2007) investigated the possibility of developing a wheat-based biorefinery to produce succinic acid. Succinic acid can be considered as a potential intermediate to produce pharmaceuticals. Besides, several value-added constituents can also be extracted from food wastes (D’Annibale et al. 2006; Bhushan et al. 2008; Silez Lopez et al. 2010). By using specific extraction methods, those constituents which possess pharmaceutical values such as essential oils and natural antioxidants can be derived from olive mill wastewaters, apple pomace and orange peel wastes. Meanwhile, Gordon and Seckbach (2012) highlighted several products and by-products derived from integrated algal biorefinery can be feed into pharmaceutical sector to converted into vitamin A, vitamin B12, antimicrobial drugs and antioxidants.

In view of abovementioned works, it is noted that it is feasible to obtain pharmaceutical intermediates from biomass feedstock. Hence, in this work, a systematic approach is presented to identify the optimum reaction pathway which converts biomass into pharmaceutical products. By producing pharmaceuticals from biomass, it may relieve several interdependent challenges that might be encountered due to the exhaustion of fossil fuel materials. This is because of most of the bulk chemicals used in producing pharmaceutical products are derivatives from fossil fuels. Besides, it can act as an alternative sustainable chemical source to produce pharmaceutical products.

Based on a comprehensive assessment of available alternatives presented in Section 2, there are multiple possible reaction pathways to convert biomass into value-added pharmaceutical products. Thus, complexity to identify the optimal pathway to produce pharmaceuticals will be increased due to an increase of degree of freedom in feedstock and products available, conversion pathways, environmental impact and safety consideration. Chemical reaction pathway synthesis method has been used as one of the important approaches to identify all possible processing routes in an integrated biorefinery (Ng et al. 2015a). It is done by conducting a comprehensive comparison on available alternatives. For example, Ng et al. (2009) introduced a hierarchical optimisation method to synthesise the potential pathways for an integrated biorefinery and identify the promising routes. Sammons et al. (2008), Bao et al. (2011) and Marvin et al. (2013) developed systematic approaches based on a flexible framework, superstructure model and supply chain network respectively for synthesising integrated biorefineries. Pham and El-Halwagi (2011) presented a method to determine the optimal pathway for a biorefinery configuration based on dynamic programming. Later, Ponce-Ortega et al. (2012) further extended this approach by proposing a disjunctive programming formulation for the design of optimal pathways for a biorefinery. Meanwhile, Murillo-Alvarado et al. (2013) introduced a systematic approach to determine the optimal reaction pathways of a biorefinery with different combination of feedstocks and products. Subsequently, Andiappan et al. (2015) (Abdelaziz et al. 2015) presented a systematic synthesis of an integrated biorefinery by incorporating multi-objective optimisation approaches. In order to identify the optimal products to be targeted in an integrated biorefinery, Ng et al. (2015b) has developed a framework that link computer-aided molecular design with the design of integrated biorefineries.

Furthermore, various objectives and criteria (e.g. economic performance, safety and health consideration, environmental impact, etc.) have been considered for the design of an integrated biorefinery. Zondervan et al. (2011) developed an optimisation model that maximise yield, minimise cost, minimise waste and fixed cost for a multi-product biorefinery system. Lately, Gong and You (2014) presented a superstructure of algal biorefinery processes for biological carbon sequestration and utilisation. Based on the superstructure, optimal design is determined via minimisation of unit carbon sequestration and utilisation cost. Recently, Abdelaziz et al. (2015) proposed a hierarchical approach for an Organocat biorefinery with a purpose of minimising its energy and mass consumption as well as total annualised cost.

While developing an integrated biorefinery model, it is discovered that there are various sources of uncertainties such as conversion rate of certain technologies, future price of products, energy and feedstock, change in climate policies as well as process scale-up. Assumptions are always made based on heuristics to fill up the information gaps where data are limited (Stuart and El-Halwagi 2013). This could greatly affect the quality of the analysis results. Sensitivity analysis is conducted in most of the existing research on the design of biorefinery to handle the uncertainties in the biorefinery system (Wang 2014). For instance, Lohrasbi et al. (2010) presented a sensitivity analysis to study the effect of plant size and transportation charges on the profitability of the process, respectively. Klein-Marcuschamer et al. (2011) conducted sensitivity analysis to address the significant areas in terms of cost savings for the ionic pre-treatment technology in a lignocellulosic ethanol biorefinery. Meanwhile, Gong and You (2014) carried out a sensitivity analysis to determine the influence of diesel price on the economic performance of algal biorefinery.

Notwithstanding the usefulness of the aforementioned works, it is realised that most of the researches have been focused on synthesise of bulk chemicals and the selection of methodologies was appropriate for that type of chemical products. The choice of value-added products to be manufactured is dependent upon the policy barriers of the biomass industry, as well as the availability of mature technology (Hart and Rajora 2009). Value-added products are only economically feasible subjected to the low capital and operating costs required or high promising revenues. Low volume yet high value products such as pharmaceutical products are yet to be implemented. The profitable manufacture of remedial products not only requires economical separation technology, but also existing market demand (Petrides et al. 2011). Pharmaceutical products derived from palm-based biomass are a new area of research which not only possessed the market demand, but also the promising high revenue even with the existing high investment needed.

Pharmaceutical product design involves specific formulation to tailor with the requirements the products’ quality profile. For the past decades, drug manufacturing has been done using pure bulk raw materials in order to comply with the Pharmaceutical Legislation of Medicinal Products for Human Use (Haleem et al. 2015). The typical production process mostly involves amalgamation of known drug substance in known dosage form and strength. Thus, the production of pharmaceutical products from palm-based biomass is a new and challenging field to be explored. Also, the pharmaceutical manufacturing industries were subjected to tight safety regulations. Plant safety is prioritised as it posed fire and explosion risk to the operating plant. Pharmaceutical industries have higher safety risks due to the handling of organic solid powders. A comprehensive analysis on assessing and retrofitting the pharmaceutical process plants have been presented by Segawa et al. (2016). The retrofitting a pharmaceutical process plant by considering the hazard assessment and life cycle analysis path flow indicators in a batch process is developed by Banimostafa et al. (2015). To determine the profitability of pharmaceuticals production, there is a need for a proper estimation method to approximate the operating cost of each pharmaceutical product. This is due to the different properties and specification requirements possessed by fine chemicals in comparison to bulk chemicals. A multi-objective optimisation-based method has been developed by Iida et al. (2015) to manufacture pharmaceutical products in a cost-effective manner by maintaining quality where as a framework to reduce energy in pharmaceutical industries has been developed by Müller et al. (2014). However, an analysis is required to investigate the uncertainty in the conversion rate of microbial fermentation technology which will aid in exploring more possibilities in this field. In addition, the other sources of uncertainties in this work are identified as the pharmaceutical products selling price and the process operating cost. This is because the price of product is controlled by the market supply and demand from time to time while the operating cost used in the optimisation model is estimated based on heuristics.

In this work, we focus on the synthesise of pharmaceutical products from palm-based biomass through optimal conversion pathway in an integrated biorefinery. This has been achieved by exploiting the potential pharmaceutical products that can be derived from palm-based biomass, by developing a new superstructure of an integrated biorefinery to produce pharmaceutical products as well as its corresponding optimisation model in terms of economic performance.


Generation of Chemical Reaction Pathway Map and Superstructure

Identification of Pharmaceutical Products

To develop systematic approach in identifying the optimum conversion pathway, the first stage is to explore potential and promising pharmaceutical products that can be derived from palm-based biomass. Palm-based biomass is mainly formed by basic molecule of lignin, cellulose and hemicellulose which are components with different composition and structure of carbon, hydrogen and oxygen. The organic bulk chemicals (intermediate products) that can be used as the intermediate products to form pharmaceutical products are identified in the next stage. These intermediates are then used as raw materials for the pharmaceutical applications. As shown in Fig. 11, chemical reaction pathway map (CRPM) is constructed for the case study to show the network and interaction between each intermediate with the biomass feedstock along with its corresponding final products. This reaction pathway network is essential to develop the superstructure of the integrated biorefinery.

Superstructure of Integrated Biorefinery Development

Figure 1 shows a superstructure is constructed which includes all the processing routes in an integrated biorefinery. As shown, the biomass feedstock b is converted to a few stages of intermediate s 1, s 2 and s 3 through the pathways q 1, q 2 and q 3 and then further processed via pathways q 4 to produce final pharmaceutical product p. The number of stages between biomass feedstock and the final pharmaceutical product are unlimited, depending on the conversion pathways found on the earlier stage. The conversion pathway that has the most number of stages is used as standard in superstructure.

Initial Screening and Data Collection for Reaction Pathways

The processing pathways are screened to filter out the immature processes by taking into consideration of product manufacturability, process maturity and feasibility. This can be further exemplified in the production of morphine from lignin as the process is not yet developed and still at research stage. Thus, these immature pathways will first be removed from analysis. Apart from that, the remaining feasible pathways will then be included in the construction of the superstructure in the subsequent step.

After identifying the potential pharmaceutical products available in the market, in formation is collected for each reaction pathway which includes the reaction conversion, plant operating cost, selling price of products and the demand of products in the market. The information collected for all the possible conversion pathways are compiled in tables to perform the comparison at ease.

Conversion, Selling Price and Market Demand of Products

The percentage conversion of various products is one of the important attributes to estimate the yield and selectivity of a specific processing route. By knowing the difference between the cost associated in producing the products and its selling prices, gross profit can be estimated for different reaction pathway. The optimisation model will select the conversion pathway that provides the highest gross margin. Besides, by obtaining the market demand of each product, production capacity of the product for that optimum processing route will not exceed the requirement of current market. Since the final products are not set at this stage, no demand requirement is set in the optimisation problem formulation.
Fig. 1

Chemical reaction pathway map (CRPM)

Plant Operating Cost

The method of estimation on plant operating cost is different for bulk and fine chemicals when the exact data from the process plants are unavailable. This is due to the different properties and specifications requirement on both chemicals. In addition, the plant operating data for fine chemicals can be very product specific and difficult to obtain. Therefore, there is a need to apply heuristics and approximate methods to get an estimate of the operating cost. Since fine chemicals are more complex and having stringent quality requirements, it is expected that the plant utilities cost especially separation/purification cost is much higher than in bulk chemicals plant (Cussler and Moggridge 2001). Multistep separation might be utilised in producing fine chemicals as well. In short, the cost of producing a fine chemical is often determined by the difficulty of purifying the commodity (Cussler and Moggridge 2001). However, apart from obtaining actual values from a commercial plant, there is no systematic method yet to determine the approximate operating cost for these fine chemicals. To obtain the scale of separating cost, heuristics-based methods such as Sherwood plots can be used. This is based on the principle that, the cost of separation depends on the dilution at which the product is produced and in the absence of more reliable data, the cost of separation can be approximated based on dilution (House et al. 2011). This heuristic gain support from a Sherwood plot (Sherwood 1959) which is a logarithmic plot covers 12 orders of magnitude of selling price.

In this work, Sherwood plot is used to approximate the correlation between cost of separating a target material and its initial concentration in a mixture. It can be observed that the slope of line in this log-log plot is negative and thus the separation cost will drop ten times if the concentration of product can be improved ten times. Therefore, the cost to separate a substance from its mixture is inversely proportional with the initial concentration of that substance. This concept has been later extended to estimate the market price and cost of production of fine chemicals when there is lack of data on the actual production cost (Cussler and Moggridge 2001; House et al. 2011). Note that the use of approximate methods is applied only as a preliminary screening stage when there is no real plant information is available.

For the pharmaceutical products that are specifically labelled in the Sherwood plot, the cost of separations is determined directly from the plot. However, there are also a few components that are not specifically labelled in the plot. For example, different types of antibiotics such as aspirin, erythromycin stearate and entacapone are included as final products in the design of superstructure. They are shown in Sherwood plot as a group of medicines which called as antibiotics. The selling prices of these antibiotics are similar to each other as well. If the separation costs of these pharmaceuticals are taken in a similar value, this will cause the model to be very sensitive to the uncertainty of operating cost. So, raw material cost is included in estimating the total operating cost of pharmaceutical product. Furthermore, it shall be highlighted that the utility cost of each commodity from the plot is known in a range of values. Since it is just a method of estimation, the actual separation cost in an industrial plant might differ from the cost obtained from Sherwood plot.

Optimisation Model for the Synthesis of Chemical Reaction Pathway

Once the potential processing routes and pharmaceutical products have been identified in the previous stages, the optimal conversion pathways to produce pharmaceuticals can be determined based on mathematical optimisation approach. In this work, the optimisation objective is set as maximise the economic performance of an integrated biorefinery. In order to develop a mathematical optimisation which can determine the optimum pathway based on the optimisation objective, an optimisation model which relates the flow of biomass to the final products through different conversion pathways is formulated. The flow rates of biomass feedstock (F 1), intermediates 1 (F 2), intermediates 2 (F 3), intermediates 3 (F 4) and products (F p ) are included in the formulation of this optimisation model.

Equation (1) below represents the splitting of biomass feedstock to all potential pathways in the q 1 level. Biomass feedstock B b Bio is converted to intermediate s 1 via pathways q 1 at a given conversion rate of R 1 to give a total production rate of intermediates 1 s 1 \( \left({T}_{s1}^{\mathrm{Inter}}\right) \) as shown in Eq. (2).
$$ {B}_b^{\mathrm{Bio}}=\sum_{q_1}{F}_1 $$
$$ {T}_{s1}^{\mathrm{Inter}}=\sum_{q1}\left({F}_1{R}_1\right) $$
Consequently, the intermediates 1 s 1 are further converted to intermediates 2 s 2 via the conversion pathways q 2. The splitting of the total production rate of intermediate \( {T}_{s1}^{\mathrm{Inter}} \) to all possible pathways q 2 with flow rate F 2 can be represented by Eq. (3).
$$ {T}_{s1}^{\mathrm{Inter}}=\sum_{q2}{F}_2 $$
The total production rate of intermediates 2 s 2 \( \left({T}_{s2}^{\mathrm{Inter}}\right) \) can be determined based on the given conversion rate of pathways q 2, R 2 via Eq. (4).
$$ {T}_{s2}^{\mathrm{Inter}}=\sum_{q2}\left({F}_2{R}_2\right) $$
Subsequently, the intermediates 2 s 2 are split to possible pathways q 3 for further conversion to intermediates 3 s 3. Later, intermediates s 3 are again to be converted to the final pharmaceutical products p via the conversion pathways q 4. The formulations for splitting the intermediates are given in Eq. (5) for intermediate 2 and Eq. (7) for intermediate 3. The total production of intermediate 3 s 3 (\( {T}_{s3}^{\mathrm{Inter}} \)) and final product p (\( {T}_p^{\mathrm{Prod}} \)) at a given conversion rate of R 3 and R 4 is shown in Eq. (6) and Eq. (8), respectively.
$$ {T}_{s2}^{\mathrm{Inter}}=\sum_{q3}{F}_3 $$
$$ {T}_{s3}^{\mathrm{Inter}}=\sum_{q3}\left({F}_3{R}_3\right) $$
$$ {T}_{s3}^{\mathrm{Inter}}=\sum_{q4}{F}_4 $$
$$ {T}_p^{\mathrm{Prod}}=\sum_{q4}\left({F}_4{R}_4\right) $$

By following Eqs. (1) to (8), the material balance of the biomass, intermediates and final pharmaceutical products can be performed.

As maximisation of economic performance is the first objective of the optimisation approach in this work, the economic performance can be defined with the following equations.
$$ \mathrm{Maximize}\ {GP}^{\mathrm{Total}} $$
$$ { G P}^{\mathrm{Total}}=\sum_p{T}_p^{\mathrm{Prod}}{G}_p^{\mathrm{Prod}}-\sum_b{B}_b^{\mathrm{Bio}}{G}_b^{\mathrm{Bio}}-\mathrm{TAOC} $$
$$ \mathrm{TAOC}=\sum_{q1}{F}_1{G}_1^{\mathrm{Opr}}+\sum_{q2}{F}_2{G}_2^{\mathrm{Opr}}+\sum_{q3}{F}_3{G}_3^{\mathrm{Opr}}+\sum_{q4}{F}_4{G}_4^{\mathrm{Opr}} $$

According to Eqs. (9) to ((11), GP Total is the gross profit of the overall integrated biorefinery configuration, TAOC is the to operating cost, \( {G}_p^{\mathrm{Prod}} \) is the selling price of p, \( {B}_b^{\mathrm{Bio}} \) is the total flow rate of biomass feedstock, \( {G}_b^{\mathrm{Bio}} \) is the cost of biomass feedstock b, \( {G}_b^{\mathrm{Opr}} \) is the operating cost for conversion of biomass b, \( {G}_{s1}^{\mathrm{Opr}} \) is the operating cost for conversion of intermediates 1 s 1, \( {G}_{s2}^{\mathrm{Opr}} \) is the operating cost for conversion of intermediates 2 s 2 and \( {G}_{s3}^{\mathrm{Opr}} \) is the operating cost for conversion of intermediates 3 s 3. In addition, constraints which applied in this model are further elaborated below.

In order to obtain the intermediates 1 s 1, biomass feedstock can be treated by using two process technologies: steam explosion and pyrolysis. The former can be applied to produce lignin, cellulose and hemicellulose simultaneously while syngas is derived from the latter. The model is subjected to mass balance constraints so that the lignocellulosic compositions will be generated together instead of syngas, and vice versa.

Flow Rate of Biomass, Intermediates and Products and Production Rate of Final Products

The flow rates of biomass, intermediates and products should not be negative values. Thus, these terms are subjected to non-negativity constraints in optimisation model as shown below.
$$ {F}_b\ge 0,{F}_{s1}\ge 0,{F}_{s2}\ge 0,{F}_{s3}\ge 0,{F}_p\ge 0 $$
In addition, the production capacity of each final product is subjected to market demand so that it will not exceed the requirement in market. An excess amount of pharmaceuticals for sale could flood the market and cause a drastic drop in selling price of the product.
$$ {F}_p\le \mathrm{market}\ \mathrm{demand} $$

Flow Rate of Biomass Feedstock

The flow rate of biomass feedstock is restricted (less than or equal to) to a certain amount. As a result of the constraints applying for the production rate of final products, as shown in section 0, reaction pathways that are not profitable might be chosen as alternatives to produce pharmaceuticals. By applying this constraint to the biomass feedstock, this allows the decision makers to check whether it is possible to obtain higher gross profit with lesser biomass feedstock.

In summary, by solving the developed mathematical model for the economic performance, the optimal conversion pathways that lead to highest profit can be determined at this stage. With the available information, mathematical model optimisation approach with different objectives such as environmental impact, sustainability of products and process safety can also be included the later stage of optimisation.

Results Analysis by Considering Uncertainties and Risky Pathways

Sensitivity Analysis

When developing the mathematical optimisation model, some of the model inputs are subject to sources of uncertainty such as the cost of raw materials, selling price of product, conversion rate of a reaction pathway and the market demand which might fluctuate along with time. In order to increase the confidence limit of the model, it must be able to cope with the variability of the system.

Sensitivity analysis is carried out to study how much the model output values are affected by the changes in model input values. The gross profit (GP) is obtained as the impact of variable to the outcome while only certain parameters which consist of uncertainties are studied. By conducting sensitivity analysis, the uncertainty of a parameter can be investigated and later provide a general assessment of model precision. Outcomes or outputs of the model are recalculated in the analysis by assuming alternative scenarios. By doing so, it is possible to determine the variables that will cause the largest outcome deviations if there is a small fluctuation in the input. In addition, the results from sensitivity analysis studies will be able to contribute to the development and improvement on the model.

Conversion Rate of Reaction Pathway

Among all parameters, conversion rate of a processing route is one of the important factors that determine the economic performance of this integrated biorefinery. In this sensitivity analysis, worst and best case scenario which is also known as extreme case analysis is assumed for each related processing route. The parameter (conversion) of each pathway is adjusted by using the most pessimistic and optimistic combination of input values which is known as extreme case analysis. The other parameters are assumed at steady state and remained as same values in base case.

Market Price of Product

The market prices of products are subjected to uncertainties like market supply and demand. To conduct the sensitivity analysis, the price of selected product is allowed to decrease gradually. The change in annual profit and optimum pathway with decreasing market price for each selected product are recorded. The graphs of profit versus product market price are plotted and will be further discussed. Note that the sensitivity analysis conducted only take into consideration of the price variation of one product at each case.

Operating Cost of Product

Operating costs of pharmaceuticals production are also subjected to uncertainty due to deficiency of data. The operating cost for each of the pathway that produces the selected products is varied from its current value. The consequence change in annual profit and optimum pathway are recorded and will be further discussed.

Feasibility Study

A number of pathways were eliminated in the earlier stage because of the lack of experimental data. The products produced from these pathways are considered as high risk products because of the uncertainty involved in production and scale-up of not established technologies. However, if these pathways produce products with high selling price, the profit margin may offset the higher production cost. To identify the high risk products with the potential to be a successful product, the profits of the eventual products produced from biomass are first calculated. The profits obtained are then compared with the market price of the high risk products. A range of operating cost for the high risk product to be competitive is also estimated in this analysis.

Alternate Solutions

Optimal design of this integrated biorefinery that will give the highest gross profit is determined after solving the mathematical model for economic performance. However, decision makers might encounter special situation which will cause the optimal solution to be infeasible when developing the integrated biorefinery. Thus, alternate solutions are generated with an attempt to provide for all contingencies.

In order to find alternate solutions, integer cuts have been applied in the model. The ‘cut’ refers to a constraint. The constraint is added to the model until the optimal basic feasible solution takes on integer value and hence a different solution is found.

In this work, the original optimal pathways are first identified in the superstructure of integrated biorefinery. Later, the flow rate of each pathway from intermediates 3, s 3, which leads to final products, p, is restricted by setting the value to zero consecutively. Thus, alternate solution is generated for each case and the result is evaluated based on economic performance and change in production rate.

The step-by-step procedures involved in the identification of optimal products and pathways from biomass are represented in Fig. 2.
Fig. 2

Procedure for the synthesis of optimal pharmaceutical products from biomass

Results and Discussions

Superstructure of Biorefinery for Pharmaceutical Production

Before developing the superstructure, screening is conducted on the conversion pathways shown in CPRM. The purpose of screening is to filter out the redundant or newer processing routes. The criteria for screening process are based on the maturity and reliability of the process where some of the pathways found in the CRPM might be still in research stage (lab-scale) and have not been commercialised in the market yet. These processes might have higher risk level in terms of risk management and economic potential compared to the other well-established and commercialised technologies. With reference to Fig. 3, the pathways compiled in the superstructure are the remaining pathways after screening process. After screening, there are 10 pharmaceutical products left to be included in the superstructure which are aspirin, paracetamol, erythromycin stearate, entacapone, formaldehyde, glucose oxidase, vitamin B12, vitamin A, penicillin and acetic acid solution.

The chemicals that are assigned as intermediates in the superstructure are mostly bulk chemicals or the chemicals that are able to sell as a product on its own. For example, methanol and benzene are the commercial bulk chemicals while vanillin can be sold as flavouring product. The superstructure and optimisation development can be simplified by choosing the correct intermediate products as each conversion path will represent a chemical plant to produce a particular intermediate or final product.

After selecting the proper intermediate products, each processing route is constructed and the number of stages of the pathway is identified accordingly. However, it is realised that different processing routes are having different number of stages. For instance, there are five stages in total to produce aspirin from biomass feedstock as this particular pathway has three intermediates to be manufactured before reaching the final product. On the other hand, there are only four stages in total for the production of glucose oxidase. So, to reduce the complexity of the structure, the conversion pathway that has the most number of stages is identified as standard in the design of superstructure. Dummy stage(s) is/are added to those conversion pathways that comprised lesser number of stages among all processing routes which are represented as blank cells in Fig. 3. By doing so, the number of stages for each conversion pathway will be the same and hence this will standardise the configuration of the superstructure. As a result, the complexity is reduced when developing the superstructural mathematical optimisation model. Since there is no reaction or interaction inside a dummy stage, it is highlighted that the conversion rate for a dummy stage is 100% and 0 in terms of operating cost of product in the optimisation model.

Besides that, with reference to the Fig. 3, dummy stage is also added at the end of a processing route which is under the same session as final product. The dummy stage represented as an alternate where the optimisation mathematical model may choose not to produce that particular final product in order to obtain higher gross profit in the model. For example, both paracetamol and acetic acid solution can be synthesised from acetic acid and the conversion pathway of producing paracetamol from acetic acid is originally selected by this model. Even though there is a drastic drop in selling price of paracetamol, it will stop producing paracetamol as final product but will not choose to produce acetic acid solution instead. Thus, in the optimisation model, acetic acid that manufactured in previous stage will proceed to the corresponding dummy stage. This is also to comply on the mass balance in the superstructure so that the unused raw material will be assigned to the dummy stage. There are a few factors that might contribute to this scenario such as low conversion rate, high operating cost or low selling price of products which will not yield a high profit.

Data Collection

The conversion and operating cost for each conversion pathway is tabulated and presented in the Appendix, Table A1. The information collected will be included in the optimisation model. Note that the prices for each product can be revised according to current market prices to conduct an up to date economic analysis. Both the tables are presented in the Appendix.
Fig. 3

Superstructure of palm-based biorefinery for pharmaceutical production

Economic Analysis for Optimisation Model

This study focuses on palm-based biomass feedstock which is known as palm kernel shell (PKS) with an annual feed flow rate of 100,000 kg. The respective composition of different components and price of the raw material is tabulated in Table A2 in the Appendix. The main objective is to synthesise an integrated biorefinery to produce pharmaceutical products with maximum economic performance. It is analysed by determining the gross profit (GP Total) obtained from the mathematical optimisation model using Eq. (10). A linear programming (LP) model is formulated in this case.

Based on the results obtained, the maximum GP Total is found to be US $403.35 million (per annum). This indicates the pathways that spawn the highest profit. The pathways selected in the model are pathways 1 (steam explosion), 2 (steam explosion), 3 (steam explosion), 5 (hydrothermal liquefaction), 8 (oxidation), 9 (hydrolysis), 10 (hydrolysis), 11 (Kolbe-Schmitt process), 12 (nitration and reduction), 17 (microbial fermentation), 18 (acidification), 24 (nitration, demethylation, knoevengel condensation), 26 (microbial fermentation) and 29 (One-Pot Synthesis). The final pharmaceutical products synthesised from this integrated biorefinery are aspirin, paracetamol, entacapone and vitamin B12. Figure 4 illustrates the synthesised configuration which represents the selected processing routes and final products.
Fig. 4

General flow diagram of synthesised integrated biorefinery

Managing Uncertainty of Conversion Rate of Processing Route

Based on the technology selected, sensitivity analysis was conducted for conversion of the processing routes. The parameter (conversion) of each pathway was adjusted by using the most pessimistic and optimistic combination of input values which is known as extreme case (worst-and-best case scenario) analysis. The results obtained from extreme case analysis are further studied and analysed.

Selected Product

Vitamin B12 is originally selected as one of the final products by the model with a conversion rate at 30%. According to Fig. 5, when the conversion rate of producing vitamin B12 from glucose is at low side (0–15%), the gross profit of the estimated from the model remained at US $62.63 million per annum. Glucose oxidase is chosen by this model as one of the final products instead of vitamin B12. This is because the processing route in producing vitamin B12 is having lesser profit during low conversion than producing glucose oxidase. When the conversion rate of glucose-vitamin B12 pathway achieved around 20%, the graph will elevate significantly. This is the turning point where vitamin B12 is chosen by the model as one of the end products and stopped producing glucose oxidase. The gross profit shows a linear increment from conversion rate of 20 till 75% and later remains almost constant until 100% conversion rate. This is because the production of vitamin B12 has reached its limit which is restricted by the constraints in this model to comply with the market demand. This indicates that it might be trivial to improve the conversion rate beyond 75% for vitamin B12 production as it will flood the market due to excess production. Once the production of vitamin B12 has reached its maximum, glucose oxidase will be chosen again as final products.
Fig. 5

Gross profit against conversion of vitamin B12

Non-selected Products

Glucose oxidase is not chosen by the model as final product. Thus, in this research, sensitivity analysis is carried out to observe whether if there is a probability to produce glucose oxidase from glucose by varying its conversion rate. With reference to Fig. 6, the gross profit remains at base value (US $403.35 million per year) from 0 until 65% conversion rate. It is then followed by a linear increment in gross profit starting from 70% conversion and the production of vitamin B12 is stopped. In other words, the conversion pathway to produce glucose oxidase from glucose can be considered as profitable if the conversion rate is above 70%. This value can be the target for researchers in order to come out with improved technology to increase the yield of glucose oxidase and to improve the economic performance of this processing route. Figure 7 illustrates the configuration which represents the production of glucose oxidase as final product.
Fig. 6

Gross profit against conversion of glucose oxidase

On the other hand, the results remain constant at base value (US $403.35 million per year) from 0 to 100% of conversion rate in penicillin, succinic acid and ethanol production. This shows that these processing routes are not as profitable as others even though conversion rates of those products have increased. The result might offer a clear direction to the decision makers so that they will be able to make a correct judgement for their investment on integrated biorefinery plant.

The results are represented in Fig. 8 to show the impact that a change in conversion rate of each pathway has on the gross profit of this model. As evidenced from the diagram, the most significant parameters that affect the gross profit of the optimisation model includes the conversion rate in production of vitamin B12 and glucose oxidase. This is due to vitamin B12 and glucose oxidase has shown wider range of profit limit acquired from variation of conversion rate in respective process. Comparatively, variation in conversion rate of acetic acid pathway shows lower range of profit limit while on the other hand penicillin, succinic acid and ethanol pathways display no effect on the model.

By plotting spider plot, it will provide an overview on the function response of different processing routes to the changes of conversion rate from base case value. With reference to Fig. 9, it can be observed that acetic acid, penicillin, succinic acid and ethanol do not show significant changes with the change in conversion rate especially in penicillin production. This has indicated that the gross profit of this model is not sensitive to the changes in conversion through these pathways. Furthermore, the gross profit remains same as in base case although the percentage change of input conversion of ethanol-penicillin pathway has already reached 700%. It can be further deduced that even though there might be a better technology to improve the yield of penicillin production in the future, this pathway will bring lesser profit compared to other processing routes. On the other hand, gross profit of the model shows the highest degree of sensitivity to the deviations of input from the original value through glucose-vitamin B12 pathway as the graph shows a steeper curve compared to the others. For glucose-glucose oxidase pathway, the gross profit shows lower degree of sensitivity compared to former as the graph remains constant at the beginning and later increases significantly when the changes in input achieved around 200%.
Fig. 7

General flow diagram for glucose oxidase production

Fig. 8

Effect of conversion rate on gross profit

Fig. 9

Changes in conversion rate of different processing routes from base case value

Managing Uncertainty of Market Price of Products

According to Eq. (10), the profit gained is proportional to the market price for the selected products. However, this is only true for the market price that is increasing. For a continuous reduction in the market price of the product, it will reduce the GP Total to a point where GP Total will not be affected by the market price of that particular product anymore. This will be discussed further below. The market price sensitivity analysis has been performed for the four chosen products, aspirin, paracetamol, erythromycin stearate and vitamin B12. The market price for each of these products is varied respectively and the results obtained are shown as below.


From the graphical representation provided in Fig. 10, it can be seen that the profit decreases linearly with the market price of aspirin until it reaches aspirin price at about US $140. This shows that even though there is a decrease of the GP Total due to lower market price of aspirin, it is still a more profitable pathway to choose than other pathways. Therefore, aspirin is still the product selected to be produced although there is a decrease in its market price.

The profit is then maintained at US $399 million per annum even for a further reduction in the aspirin market price. This is because the huge decreased in aspirin market price has caused the initial optimum pathways to change. For the aspirin price lower than US $140, instead of producing aspirin, it is found that producing higher amount of erythromycin stearate is more profitable. Hence, the optimal products for aspirin market price lower than US $140 are changed to produce paracetamol, erythromycin stearate and vitamin B12 only where the production rate of erythromycin stearate is increased from 584.50 kg (per annum) to 857.47 kg (per annum) while the production rate of the rest remains the same.
Fig. 10

Aspirin price sensitivity analysis


As can be observed in Fig. 11, paracetamol is produced from two different intermediates which are lignin and hemicellulose via different pathways. From Fig. 11, it shows that when the price of paracetamol decreases, profit also decreases. At a point where market price of paracetamol has fall until US $140, the graph starts to deviate from its linearity. This is because the pathway of lignin-producing paracetamol earlier on has been shifted to produce more erythromycin stearate which gives higher profit than producing paracetamol at low selling price. However, the falling price of paracetamol is still affecting the annual profit because paracetamol is still being produced from another intermediate, hemicellulose.

Note that paracetamol is still being produced from hemicellulose instead of acetic acid solution despite its price has dropped to US $0. This is mainly due to the low operating cost of the conversion process of paracetamol compare to acetic acid solution which has outweighed the profit obtained from producing acetic acid solution. However, the production of these two products should be forgone to achieve optimal profit. Thus, empty boxes have been added in the superstructure of integrated biorefinery when taken into consideration of this scenario.
Fig. 11

Paracetamol price sensitivity analysis

Erythromycin Stearate

In Fig. 12, it illustrates a proportional relationship between profit and market price of erythromycin stearate. Besides that, it also shows that the optimal pathways generated are independent of the market price of erythromycin stearate. When the market price of erythromycin stearate is zero, the annual profit generated is US 4.01 million. However, the profit dropped to US 381 million (per annum) once the production of erythromycin stearate is ceased. Hence, it can be concluded that the optimisation model will always choose to produce erythromycin stearate regardless of its selling price due to its extremely low operating cost. However, this pathway should be omitted and send for disposal when the operating cost of its conversion pathway is exceeding the market price of erythromycin stearate.
Fig. 12

Erythromycin stearate price sensitivity analysis

Vitamin B12

This scenario is similar to the aspirin case. Figure 13 shows a plateau at annual profit of US $64 million which indicates a changed of optimum pathways. When the price of vitamin B12 dropped to US $110,000, it is found that producing glucose oxidase will be more profitable than producing vitamin B12. Therefore, further decrease of the market price of vitamin B12 will no longer affect the annual profit as the optimum pathway has shifted from vitamin B12 to glucose oxidase.
Fig. 13

Vitamin B12 price sensitivity analysis

To summarise, in order to keep these products as possible pathways and make profit, the market prices of these products must be more than a certain minimum value. These minimum market prices of the products can be determined through this analysis. However, for the product that have extremely low operating cost such as erythromycin stearate, its market price can only be as low as its operating cost.

The minimum market prices of aspirin, paracetamol, erythromycin stearate and vitamin B12 in this analysis are found to be US $140/kg, US $140/kg, US $3610/kg and US $110,000/kg, respectively. If the actual market prices of these products are lower than the given minimum market price, the optimal pathway of the superstructure will be altered. The percentage of allowable price drop from original market price for aspirin, paracetamol, erythromycin stearate and vitamin B12 are 77, 66, 18 and 38%, respectively.

Operating Cost of Final Product

As the operating costs for pharmaceuticals production are estimated by the heuristics developed, the actual operating costs are more likely to be deviated. In order to study the relationship between the annual profit and the change of operating cost, a range of operating cost and its respect profit are tabulated in Table A3 in the Appendix.

According to the results obtained from the sensitivity study, all annual profit decrease linearly when the corresponding operating cost increases except for the conversion pathway of amino phenol to paracetamol. Initially paracetamol is being produced mainly from amino phenol and acetic acid. However, this production pathway of paracetamol changes when the operating cost of amino phenol to paracetamol increases 50% from current operating cost. Due to the increase in operating cost, the optimisation model chooses to abandon the conversion pathway of amino phenol to paracetamol and adopts another pathway. Paracetamol is now also being produced by chlorobenzene and acetic acid.

Feasibility Study for Potential Pathways

Some of the products have been omitted at the screening stage due to insufficient information of the conversion pathway or immaturity of the technology. However, those products may have high profit margin and can be competitive to the selected products. Thus, it is good to revisit those pathways to identify the potential ones.

The potential risky pathways that will be revised in this analysis are shown as follows:
  • Morphine from vanillin

  • Lomotil from lignin

  • Phenanthrene from cellulose

  • Xylitol natural sweetener from C5 and C6 sugars

Note that these pathways will be compared with the eventual product that produced from the same intermediates. For instance, morphine from lignin is comparing with erythromycin stearate which also derived from lignin. The market prices of those revised products are obtained from literature and tabulated in Table A4 in the Appendix.

It is estimated that, erythromycin stearate from lignin has a profit of US $1200/kg. The selling price of morphine (US $45,500/kg) is much higher than the profit obtained from selling erythromycin stearate. Therefore, it can be deduced that if morphine is produced from lignin at a cost not more than US $44,300/kg, it will be a competitive product to be produced as illustrated in Fig. 14.
Fig. 14

Graphical comparison for erythromycin stearate and morphine

As to know whether lomotil is a potential feasible product to be produced, profit of all the eventual products from the same intermediate (lignin) is calculated and compared as shown in Fig. 15. The profits obtained from lignin to aspirin, paracetamol and erythromycin stearate are US $425/kg, US $283/kg and US $1083/kg, respectively. Since the market price of lomotil is higher than each of the profit gained from the product of same intermediate, lomotil is a promising product to be produced. For lomotil to be competitive in market, the maximum operating cost for lomotil production should be not more than US $16,000/kg.
Fig. 15

Lomotil, aspirin, paracetamol and erythromycin stearate pathways comparison

Conversely, the profit resulted from producing vitamin B12 is US $79,000/kg, which is higher than the selling price of phenanthrene. In addition, xylitol natural sweetener also has much lower market price than the profit obtained from converting C5, C6 sugars to paracetamol. Therefore, it can be concluded that phenanthrene and C5, C6 sugars cannot be competitive no matter how low their operating costs are.

Alternate Solutions

The other possible products that can be produced from biomass according to the general flow diagram shown in Fig. 4 are analysed. There are five alternate optimal solutions obtained from this optimisation model and the profit of each scenario is tabulated in Table 1. Each alternate solution is compared with the original feasible solution to evaluate the difference in terms of economic performance. The purpose of finding alternate solutions is to enable the decision makers to be prepared for different situations. For instance, it is found out that the operating cost of a product is significantly higher in commercial scale than base case value. Or else there might be a time of scarcity for certain raw material. Therefore, alternate solution which shows minor difference with profit from base case can be taken into consideration to replace the original pathway in the integrated biorefinery.
Table 1

Alternate optimal solutions




Profit (million USD/year)

Difference with profit from base case (%)



Sodium salicylate





Amino phenol




Erythromycin stearate





Vitamin B12






Acetic acid



For case 1, when the production of aspirin is restricted, production rate of erythromycin stearate will increase. This has caused a minor loss in profit as manufacturing erythromycin stearate brings lesser revenue in the integrated biorefinery. While for case 2, production rate of erythromycin stearate increases but the profit drops slightly if paracetamol is not produced from amino phenol. This is because paracetamol is produced from two processing routes where the pathway through acetic acid is not as profitable as from amino phenol.

Besides that, in case 3, when the production of erythromycin stearate is restricted, entacapone will be selected by the model as one of the final products with 5.1% profit loss. In case 5, the demand of paracetamol can be fulfilled by producing it from amino phenol instead of acetic acid, still there is a minor profit loss in integrated biorefinery. On the other hand, by stopping the production of vitamin B12 in case 4, the biorefinery might encounter 85% profit loss which is an unfavourable situation to a decision maker. This is because the selling price of vitamin B12 is a lot higher than glucose oxidase and low conversion rate for glucose oxidase production makes it impossible to compensate the loss in profit. In summary, cases 1, 2, 3 and 5 which show a small deviation from original profit have indicated that these alternative solutions are feasible to replace the original pathways.


In this research, a systematic approach was developed to synthesise pharmaceutical products from biomass through optimal processing route. Different potential pharmaceutical products that can be derived from palm-based biomass were included in the design of superstructure of integrated biorefinery. To evaluate these processing routes in the integrated biorefinery, a mathematical optimisation approach was introduced to synthesise the integrated biorefinery. Hence, the optimum conversion pathways that convert biomass into pharmaceutical products were identified based on economic performance. Apart from that, analyses were carried out to investigate the impact of changes in conversion rate, operating cost and market price to the economic performance of this integrated biorefinery, respectively. Also, feasibility study was carried out to identify those potential pathways among the eliminated ones in terms of economic performance. Alternate solutions were determined to replace the original pathways during different situations. In a nutshell, the optimised methodology proposed for converting palm-based biomass into pharmaceutical product design by considering the uncertainties is achieved in this work.

In addition, there are some potential areas of improvement in this research. A systematic approach on establishing appropriate weighing factors is suggested on the redundant or newer processing routes that have been screened out in earlier stage in order to form a risk level decision matrix. By doing so, the process technologies that have been filtered out can be quantified in order to determine the feasibility of those conversion pathways. There are a few aspects that can be taken into consideration in developing a decision matrix such as economic potential, customer acceptability, environmental friendliness, technology maturity and reliability. It is to be noted that the developed approach is to be used as a screening tool for identifying the most promising pathways before conducting a rigorous analysis on selected pathways. In the final decision making stage, the environmental, social and safety aspects must be covered in detail. A more rigorous economic analysis must also be performed at that stage. Future work will be also directed towards extending the sensitivity analysis on conversion rate by considering the variation of operating cost. This model can be updated if an improved process with better conversion is developed in the future.



This work was supported by the Ministry of Science, Technology and Innovation (MOSTI) Malaysia under Grant no. 06-02-12-SF0224.

Compliance with Ethical Standards

Conflict of Interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.

Supplementary material

41660_2017_10_MOESM1_ESM.docx (28 kb)
ESM 1 (DOCX 28 kb)


  1. Abdelaziz O, Gadalla M, El-Halwagi M, Ashour F (2015) A hierarchical approach for the design improvements of an Organocat biorefinery. Bioresour Technol 181:321–329CrossRefGoogle Scholar
  2. Andiappan V, Ng D, Bandyopadhyay S (2014) Synthesis of biomass-based trigeneration systems with uncertainties. Ind Eng Chem Res 53(46):18016–18028CrossRefGoogle Scholar
  3. Andiappan V, Ko A, Lau V, Ng L, Ng R, Chemmangattuvalappil N, Ng D (2015) Synthesis of sustainable integrated biorefinery via reaction pathway synthesis: economic, incremental environmental burden and energy assessment with multiobjective optimization. AICHE J 61(1):132–146CrossRefGoogle Scholar
  4. Banimostafa A, Papadokonstantakis S, Hungerbühler K (2015) Retrofit design of a pharmaceutical batch process considering “green chemistry and engineering” principles. AICHE J 61(10):3423–3440CrossRefGoogle Scholar
  5. Bao B, Ng D, Tay D, Jiménez-Gutiérrez A, El-Halwagi M (2011) A shortcut method for the preliminary synthesis of process-technology pathways: an optimization approach and application for the conceptual design of integrated biorefineries. Comput Chem Eng 35(8):1374–1383CrossRefGoogle Scholar
  6. Bhushan S, Kalia K, Sharma M, Singh B, Ahuja P (2008) Processing of apple pomace for bioactive molecules. Crit Rev Biotechnol 28(4):285–296CrossRefGoogle Scholar
  7. Cussler E, Moggridge G (2001) Chemical product design. Cambridge University Press, CambridgeGoogle Scholar
  8. D’Annibale A, Sermanni G, Federici F, Petruccioli M (2006) Olive-mill wastewaters: a promising substrate for microbial lipase production. Bioresour Technol 97(15):1828–1833CrossRefGoogle Scholar
  9. Du C, Lin S, Koutinas A, Wang R, Webb C (2007) Succinic acid production from wheat using a biorefining strategy. Appl Microbiol Biotechnol 76(6):1263–1270CrossRefGoogle Scholar
  10. Fernando S, Adhikari S, Chandrapal C, Murali N (2006) Biorefineries: current status, challenges, and future direction. Energy Fuel 20(4):1727–1737CrossRefGoogle Scholar
  11. Gnansounou E, Dauriat A (2010) Techno-economic analysis of lignocellulosic ethanol: a review. Bioresour Technol 101(13):4980–4991CrossRefGoogle Scholar
  12. Gong J, You F (2014) Optimal design and synthesis of algal biorefinery processes for biological carbon sequestration and utilization with zero direct greenhouse gas emissions: MINLP model and global optimization algorithm. Ind Eng Chem Res 53(4):1563–1579CrossRefGoogle Scholar
  13. Gordon R, Seckbach J (2012) The science of algal fuels. Springer, Dordrecht, NetherlandCrossRefGoogle Scholar
  14. Haleem R, Salem M, Fatahallah F, Abdelfattah L (2015) Quality in the pharmaceutical industry—a literature review. Saudi Pharmaceutical Journal 23(5):463–469CrossRefGoogle Scholar
  15. Harisha AS, Nayak SP, Pavan MS et al (2015) A new synthesis of Entacapone and report on related studies. J Chem Sci 127(11):1977–1991CrossRefGoogle Scholar
  16. Hart CA, Rajora ML (2009) Overcoming institutional barriers to biomass power in China and India. Sustainable Development Law and Policy 9(3):26–31 64-65Google Scholar
  17. House K, Baclig A, Ranjan M, van Nierop E, Wilcox J, Herzog H (2011) Economic and energetic analysis of capturing CO2 from ambient air. Proc Natl Acad Sci 108(51):20428–20433CrossRefGoogle Scholar
  18. Huber GW, Corma A (2007) Synergies between bio- and oil refineries for the production of fuels from biomass. Angew Chemie Int Ed 2007(46):7184–7201CrossRefGoogle Scholar
  19. Iida T, Ota K, Sasozaki T, Sugiyama H (2015) Multiobjective design method for the use processes of pharmaceutical excipients considering quality and cost-effectiveness. J Pharm Innov 10(4):313–323CrossRefGoogle Scholar
  20. Kimble M, Pasdeloup M, Spencer C (2008) Sustainable bioenergy development in UEMOA member countries. West African Economic and Monetary Union, Ouagadougou, Burkina FasoGoogle Scholar
  21. Klein-Marcuschamer D, Simmons B, Blanch H (2011) Techno-economic analysis of a lignocellulosic ethanol biorefinery with ionic liquid pre-treatment. Biofuels Bioprod Biorefin 5(5):562–569CrossRefGoogle Scholar
  22. Lohrasbi M, Pourbafrani M, Niklasson C, Taherzadeh M (2010) Process design and economic analysis of a citrus waste biorefinery with biofuels and limonene as products. Bioresour Technol 101(19):7382–7388CrossRefGoogle Scholar
  23. Marvin W, Schmidt L, Daoutidis P (2013) Biorefinery location and technology selection through supply chain optimization. Ind Eng Chem Res 52(9):3192–3208CrossRefGoogle Scholar
  24. Mohan T, El-Halwagi M (2006) An algebraic targeting approach for effective utilization of biomass in combined heat and power systems through process integration. Clean Techn Environ Policy 9(1):13–25CrossRefGoogle Scholar
  25. Müller G, Sugiyama H, Stocker S, Schmidt R (2014) Reducing energy consumption in pharmaceutical production: framework and case study. J Pharm Innov 9:212–226CrossRefGoogle Scholar
  26. Murillo-Alvarado P, Ponce-Ortega J, Serna-Gonzalez M, Castro-Montoya A, El-Halwagi M (2013) Optimization of pathways for biorefineries involving the selection of feedstocks, products, and processing steps. Ind Eng Chem Res 52(14):5177–5190CrossRefGoogle Scholar
  27. Naik S, Goud V, Rout P, Dalai A (2010) Production of first and second generation biofuels: a comprehensive review. Renew Sust Energ Rev 14(2):578–597CrossRefGoogle Scholar
  28. Ng D, Pham V, El-Halwagi M, Jimenez-Gutierrez A, Spriggs H (2009) A hierarchical approach to the synthesis and analysis of integrated biorefineries. Design for Energy and the Environment: Proceedings of Seventh International Conference on Foundations of Computer-Aided Process Design 1(8):425–432CrossRefGoogle Scholar
  29. Ng L, Andiappan V, Chemmangattuvalappil N, Ng D (2015a) Novel methodology for the synthesis of optimal biochemicals in integrated biorefineries via inverse design techniques. Ind Eng Chem Res 54(21):5722–5735CrossRefGoogle Scholar
  30. Ng L, Andiappan V, Chemmangattuvalappil N, Ng D (2015b) A systematic methodology for optimal mixture design in an integrated biorefinery. Ind Eng Chem Res 81:288–309Google Scholar
  31. Petrides D, Siletti C, Jimenez J, Psathas P, Mannion Y (2011) Optimizing the design and operation of fill-finish facilities using process simulation and scheduling tools. Pharm Eng 2:1–10Google Scholar
  32. Pham V, El-Halwagi M (2011) Process synthesis and optimization of biorefinery configurations. AICHE J 58(4):1212–1221CrossRefGoogle Scholar
  33. Pokoo-Aikins G, Nadim A, El-Halwagi M, Mahalec V (2009) Design and analysis of biodiesel production from algae grown through carbon sequestration. Clean Techn Environ Policy 12(3):239–254CrossRefGoogle Scholar
  34. Ponce-Ortega J, Pham V, El-Halwagi M, El-Baz A (2012) A disjunctive programming formulation for the optimal design of biorefinery configurations. Ind Eng Chem Res 51(8):3381–3400CrossRefGoogle Scholar
  35. Ravaghi-Ardebili Z, Manenti F (2015) Unified modelling and feasibility study of novel green pathway of biomass to methanol/dimethylether. Appl Energy 145:278–294CrossRefGoogle Scholar
  36. Sammons N, Yuan W, Eden M, Aksoy B, Cullinan H (2008) Optimal biorefinery product allocation by combining process and economic modeling. Chem Eng Res Des 86(7):800–808CrossRefGoogle Scholar
  37. Segawa A, Yoshikawa S, Toyama T, Nakanishi H, Kikuchi-Uehara E, Hirao M, Sugiyama H (2016) Method for reducing environmental, health, and safety risks in active pharmaceutical ingredient manufacturing based on multiobjective evaluation. Process Saf Environ Prot 104:304–313CrossRefGoogle Scholar
  38. Sherwood T (1959) Mass transfer between phases. Pennsylvania State University, University Park, TexasGoogle Scholar
  39. Silez Lopez A, Li Q, Thompson I (2010) Biorefinery of waste orange peel. Natl Center Biotechnol Inf 30(1):63–69Google Scholar
  40. Stuart P, El-Halwagi M (2013) Integrated biorefineries. CRC Press, Boca RatonGoogle Scholar
  41. Wang L (2014) Sustainable bioenergy production. CRC Press, Taylor and Francis, Boca RatonCrossRefGoogle Scholar
  42. Zondervan E, Nawaz M, de Haan A, Woodley J, Gani R (2011) Optimal design of a multi-product biorefinery system. Comput Chem Eng 35(9):1752–1766CrossRefGoogle Scholar

Copyright information

© Springer Nature Singapore Pte Ltd. 2017

Authors and Affiliations

  • Shi Yee Ng
    • 1
  • Sze Ying Ong
    • 1
  • Yee Yin Ng
    • 1
  • Alexander H. B. Liew
    • 1
  • Denny K. S. Ng
    • 1
  • Nishanth G. Chemmangattuvalappil
    • 1
  1. 1.Department of Chemical and Environmental Engineering/Centre of Sustainable Palm Oil Research (CESPOR)The University of Nottingham Malaysia CampusSemenyihMalaysia

Personalised recommendations