While it is certainly plausible that an industrial hydrolysis step can yield 100% conversion, this small decrease in COGS may not be worth the added difficulty of extending the hydrolysis time or effort. Thus, it is suggested that a 94 percent conversion may a reasonable operating condition to aim for.Extraction of resveratrol in plants has been demonstrated using variety of techniques, such as UAE and Soxhlet extraction. However, as mentioned above, certain bio-processing values such as percentage recovery for resveratrol extraction from Japanese knotweed using UAE is not specified in literature. Thus, leading to us using a percent recovery value derived from literature on ultrasonic assisted extraction of resveratrol from grape stems in our base case model, a value of 78.8%. Interestingly, the same authors describe a significant increase in percent recovery when the plant biomass undergoes a subsequent agitation with fresh solvent, increasing the recovery up to 96.7%10. In efforts to assess the relationship between extraction recovery and CAPEX and COGS, the extraction process was simulated to undergo two agitation steps with the introduction of fresh ethanol , while varying the percent recovery by increments of two from 90% to 100%. Figure 4.10 demonstratesa graphical representation of the relationship between CAPEX, COGS, garden grow bags and percent recovery during the extraction step.During the extraction efficiency sensitivity analysis, an immediate outlier was determined.
When the extraction efficiency was set to a 94% recovery, the CAPEX saw a reduction in price of $1.7 million when compared to the 92% recovery. Interestingly, the CAPEX only dropped an average of $60 thousand among other increments. Upon further inspection, the drop in CAPEX was attributed to multiple factors which effect the DFC and the reduction in total plant direct costs . First, 19% of the $1.7 million dollar reduction is directly related to the removal of one ultrasonic assisted extraction unit needed for processing. With the increase in extraction efficiency, the mass of knotweed required for processing decreases , subsequently requiring less equipment and piping needed to installation . The simulation model was able to recognize that 9 UAEs were enough to process the incoming mass instead of the 10 initially required in the first initial scenarios , 90% and 92% recovery. Along with the decrease in listed and unlisted equipment, the simulation model also reduced the costs associated with installation, process piping, instrumentation, insulation, electrical, building, yard improvements, and auxiliary facilities, leading to an additional decrease in price . A large portion of the CAPEX reduction is credited to the decrease in total plant indirect costs , which incorporates the engineering and construction costs associated to the design and construction of the process. The engineering and construction costs are calculated as a factor of the total direct cost, 25% and 35% of the DC respectively. The remaining price decrease is attributed to the reduction in contractors fee , working capital and startup costs. Overall, there is a downward trend and decrease in the CAPEX and COGS when extraction efficiency was optimized.
While an extraction efficiency of 94% seems to be the best percentage to aim for, the efficiency of resveratrol being extracted from Japanese knotweed during a UAE step still requires to be confirmed in a laboratory scale prior to scaling to large scale, as values may vary from those simulated here.As certain technologies improve, and more pilot and production scale data become available, certain bio-processes parameters like resin exchange frequency, enzymatic hydrolysis conversion, and extraction efficiency can be tuned. An alternative approach to pursuing these technologies might be to redesign current facilities using current technology in attempts to reduce CAPEX and/or COGS. Here, acknowledging that the amount of ethanol needed in our process is the bottle neck, we performed a scenario analysis where we simulate the addition of an ethanol recovery unit to the current base case model. In the base case model, ethanol is not being recycled and is disposed of immediately after being used in the extraction unit and the adsorption vessel . In efforts to reduce the overall cost of ethanol, an additional plate and frame filter was introduced directly after the initial plate and frame filter to increase the separation between ethanol and plant biomass during the separation process. The ethanol recovered was sent to a holding vessel for future processing. Additionally, the ethanol used to elute impurities from the NKA-II resin in the adsorption vessel was captured and sent to the same holding vessel. The ethanol within the holding vessel was then transferred into a distillation column where ethanol is distilled and separated from any other impurities recovered.
The vapor is sent to a condenser where the exiting liquid ethanol is then recycled to the ultrasonic assisted extractors for reprocessing. A comparison of CAPEX and COGS for a facility with and without the ethanol recovery equipment is shown in Figure 4.13 and Figure 4.14, respectively.When comparing the difference in CAPEX between both scenarios, the model recycling ethanol was determined to be $6.8 million more costly than the base case, for a total cost of $51.5 million. The justification for this arises from the additional equipment needed to ensure sufficient recovery and processing of ethanol needed for recycling within the process, i.e., a plate and frame filter, a distillation unit, holding vessel, and a condenser. While CAPEX is higher for the new proposed facility, both the cost of ethanol and COGS decrease. The base case model utilizes about over 6 million kg of ethanol to process 100 MT of resveratrol. The proposed facility suggests a reduction of 70% of total ethanol used, only utilizing 2.1 million kg of ethanol per year needed for processing. The cost in ethanol in the base case was determined to be over $4 million whereas when the recovery unit was added, the cost of ethanol was reduced to $1.3 million, a decrease of 67% which aligns with the recovery percentage calculated. This reduction in ethanol is reflected in the COGS of the two facilities, shown graphically in Figure 4.14.The model suggests that the COGS for the base case simulation is about $149.5/kg including depreciation and $111.5/kg Rsv without including depreciation. Here, it is shown that the addition of a recycling stream can reduce the COGS down to $141.4/kg resveratrol including depreciation and $96.8/kg Rsv without including depreciation. Although the reduction in ethanol was significant, the overall COGS did not reduce in a similar fashion. The difference in price is $7.5 for COGS including depreciation and $ 14.2 for COGS not including depreciation. One reason the scenario model did not reduce COGS further is because of the increase in labor attributed to the new pieces of equipment needed for recycling ethanol. In the scenario simulation, labor costs increased $400 thousand to $1.4 million from $1.0 million, a 26% increase. In this analysis, labor time allocated for certain operations used across the model remained constant , however, tomato grow bags the additional labor time was allocated to the distillation, condenser, plate and frame filter, and extractor unit. Furthermore, the cost for water increased over $10 thousand/year in the ethanol recovery model since fresh water was being added into the recycling stream to reduce the concentration of the recycle stream to create optimal conditions for extraction to be performed. One method the COGS can further be reduced is if water was also recovered leaving the distillation unit and recycled. The authors highly recommend implementation of such recycling unit for large scale processing.One scenario analysis which may be of inTherest is the relationship between COGS, CAPEX, and annual production amount. Through research and personal communications with resveratrol manufacturers in China, it has been understood that 100 MT may account for 1/3 of the global resveratrol production market, dominating production over current resveratrol manufacturers.
To assess the impact the production amount with COGS and CAPEX, we simulated bio-manufacturing facilities operating under similar conditions as the base case but now capable of producing amounts ranging from 25 to 200 MT. A new SuperPro file was designed for each annual production model where equipment was resized appropriately. Figure 4.15 demonstrates the relationship between production amount to CAPEX and COGS.As expected, the largest difference in CAPEX and COGS price arose between the base case and the models producing the lowest and highest amount of resveratrol, 25 and 100 MT respectively. The average difference in CAPEX between each model and the base case was about $23 million, with the exact values being $25.6 million and $71.0 million for the 25 and 200 MT simulations, respectively. It was determined that increasing the production amount led to an increase in the annual operating costs. However, the larger production amount resulted in a lower COGS value. These results were anticipated as the notion of economy of scales is demonstrated, . When an economic analysis was performed on the simulation producing 200 MT, the COGS was determined to be $124/kg, a reduction of $25 from the base case. The inverse is accurately represented within our results as well. While the CAPEX and OPEX for a bio-manufacturing facility producing 25 MT of resveratrol did decrease, the COGS increased significantly, resulting in a COGS of $311/kg resveratrol, an increase of $162 from the base case. The COGS for all the production amount scenarios are shown below in Figure 4.16. One assumption made during this scenario analysis was that the price of knotweed rhizomes remains a constant price of $0.19/kg. This value was calculated as the cost for the base case when 7.3 million kg of knotweed rhizomes was needed. This assumption would be valid if the quantity of knotweed being harvested annually remained the same. However, there would be a large difference when altering production amount since the mass of knotweed varied with the production amount. Specifically, when the biomanufacturing facility was modeled to produce 25 MT, only 1,415 kg of knotweed rhizomes was required per batch. In the model where the aim was to produce 200 MT of resveratrol, over 11,000 kg of rhizomes were needed per batch. One justification for not increasing the price of knotweed rhizomes is acknowledging that the concentration of resveratrol per knotweed rhizome varies on environmental factors and a larger batch may be needed to offset the low concentrations per plant. The reverse can also be said for the 200 MT model. If the concentration of resveratrol from the base case were doubled to 1.0 mg/g, the number of plants required to produce 200 MT originally would be halved and the current base case would serve as an appropriate model.As an evaluation of the sustainability of our proposed facility, we conducted two environmental assessments on the base case scenario. Following methods listed by Budzinski et al., and by Biwer and Heinzle, the process mass intensity and Environmental, Health, and Safety Assessment were calculated, respectively. The PMI compares how much of a certain component of the mass balance is consumed or formed per unit amount of final product, depicting the mass or material efficiency of the process1 . A completely efficient and ideal process has a PMI value close to 1 indicating an efficient process. A large PMI value indicates a large input is required for small amounts of yield. Budzinski’s method of calculating the PMI fails to include inputs used for cleaning during the cleaning-in-place operation. In this report, the calculation of PMI will include cleaning inputs since exclusion of these inputs may result in an inaccurate environmental analysis. After performing the PMI procedure with the cleaning inputs, the proposed process was found to have an overall PMI of 529 kg inputs/kg resveratrol. A breakdown of the PMI is shown graphically in Figure 5.1. As expected for plant bio-processing, the PMI value is dominated by water consumption, which accounts for 72% of the overall PMI. Water is used throughout the process to wash the incoming plants, create slurry solutions, as a component for extraction, and for rinsing and cleaning certain units before each operation. The next largest contributor is the Japanese knotweed rhizomes. The plant biomass needed to produce 100 MT makes up about 14% of the overall PMI by itself. The only consumable listed within the simulation is the resin used for purification within the adsorption vessel.