Design of an Industrial Process for Enzymatic Cannabidiol Conversion
No Thumbnail Available
Date
2024
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of the Witwatersrand, Johannesburg
Abstract
This work aims to model a theoretical enzymatic bioreactor and all the necessary surrounding processes required to facilitate the bioremediation of THC into CBD, to produce a CBD product with THC levels below the legal concentration limits (0.001%). The primary purpose is to explore whether further research into the potential biochemical remediation of THC into CBD would be worth pursuing in terms of both functionality and profitability within the CBD industry. Two primary process designs were modelled using SuperPro Designer, one producing a CBD isolate, and another producing a full spectrum CBD blend containing other cannabinoids beyond CBD, as well as other compounds like flavonoids and terpenes. The CBD isolate model is composed of four parts: the extraction of the crude oil (including the pre-extraction process); the upstream processing of the oil; the reaction of THC into CBD; and the downstream processing of the oil. The full spectrum CBD model is similarly structured but with a different fourth process stage (downstream processing). Cultivating one’s own cannabis was calculated to be more economical than purchasing it from a third-party supplier, and thus a drip-based irrigation system of 1.2108 hectares was used, requiring capital costs of $ 24 641.90 and a yearly cultivation cost of $ 7629.76. Both processes begin with milling to increase the surface area of the cannabis, followed by passing through two consecutive cold ethanol mixer-settler extraction units. Next, the oil-plant matter mixture passes through a plate-and-frame filtering system and then a decarboxylation oven, which will convert the cannabinoids into their neutral forms, producing CBD from CBDA and THC from THCA, and releasing CO2 as a co-product. The oil then passes into a PFR, where CLEAs catalyse the reaction of THC into CBD. Due to the theoretical nature of the as-yet-unknown enzyme, conversion was assumed to be 85 %, where 37.46 kg of the enzyme was calculated to be required per year, assuming replacement is required after seven days of operation. The possibility of either producing or purchasing the enzyme was considered, but producing the enzyme worked out more economically viable, at a yearly cost of $ 2.84. This is where the full spectrum process halts. In the isolate process, the oil is then mixed with an acetonitrile-TBME stream before entering the CPC, alongside a heptane stream. Most of the CBD passes into the heptane, which moves iv to a distillation column after exiting the CPC process, with CBD isolate emerging from the bottom stream. The acetonitrile-TBME stream exiting the CPC will contain other cannabinoids and remaining cannabis compounds and will flow into a separate distillation. The bottom stream of the distillation column provides other valuable cannabinoid isolates, forming ancillary products (CBG, CBN, and THC isolates). In the full spectrum CBD model, the oil flows directly into a series of three consecutive distillation columns upon exiting the PFR, designed to reduce the THC concentrations in the oil to acceptable levels. The CBD oil emerging from both processes must then be incorporated into an MCT carrier oil. CBD isolates were assumed to be sold for $ 39.00 per 30 ml, with each unit containing 600 mg of CBD. Full spectrum CBD oil was assumed to be sold for $ 40.00 per 30 ml, where every 30 ml contains 1500 mg of cannabis oil. Once the costs of the MCT oil were deducted from the theoretical revenue values, the net revenue values came to $ 64 089.93 per kilogram for CBD isolate and $ 26 302.64 per kilogram for the full spectrum CBD oil. Collectively, the ancillary cannabinoid products (CBG, CBN, and THC) yielded a net revenue value (less the cost of the required MCT oil) of $ 17749.48 /kg. CBD isolate was produced at a rate of 637 kg/year, at 99.47 % purity, with ancillary cannabinoid products being produced at 494 kg/year by the isolate process. The full spectrum CBD blend was produced at a rate of 656 kg/year and did not contain any solvent residues. The isolate process was found to have a gross margin of 83.86%, an ROI of 101.58%, a payback time of 0.98 years, and an NPV of $ 190 458 000. The full spectrum blend has a gross margin of 50.89%, an ROI of 21.75%, a payback time of 4.60 years, and an NPV of $ 14 199 000. Thus, the isolate process was deemed the more economically viable of the two processes. An additional CBD isolate design involving supercritical CO2 extraction was also modelled for comparison. In this variation, the cannabis buds undergo milling before passing through the supercritical CO2 extraction unit. The CO2-ethanol solvent feed enters a CO2 storage unit wherein is pressurised to achieve supercritical conditions before entering the extraction unit alongside the cannabis stream. The bottom stream from the extraction unit then passes into a plate-and-frame filtration system, which removes the plant matter from the stream; the recovered cannabis oil is then reunited with the top stream from the extraction unit. The combined oil stream then undergoes winterisation, in 24-hour cycles, before moving into a distillation column which removes any remaining solvent. The bottom stream from the column then enters the decarboxylation oven; the remainder of the process continues in the same manner as the original, cold ethanol extraction isolate process. The CO2 extraction process produced CBD isolate at a rate of 600.54 kg/year. CBD purity of 99.29 % was achieved. An economic analysis produced project indices of a gross margin of 84.07%, an ROI of 95.28%, a payback time of 1.05 years, and an NPV of $ 173 404 000. Thus, with gross margin being the sole exception, all project indices indicate the cold ethanol process being the process with greater potential profitability to produce CBD isolates. Because the isolate process proved the most profitable of the alternatives, its potential profitability when scaled up to industrial size was also assessed. The process feed rate was increased to 79 200 kg of cannabis buds per year, solvent input streams were proportionally scaled up, and several equipment units were multiplied as required. Additionally, the quantity of the enzyme required for catalysing the reaction was recalculated based on the increased plant material in the process, coming to a yearly mass of 576.25 kg, for $ 345 750. The scaled-up process produced a CBD isolate product with a purity of 99.47% and a production rate equivalent to 9800 kg per year and the ancillary cannabinoid product at a rate of 7587 kg per year. The NPV of the scaled-up process came to $ 3.99 billion and a gross margin of 99 % was achieved, with an ROI of 1340 % and a payback time of 0.07 years. Therefore, from the simulated model and the economic analyses, the production of CBD oils using THC-to-CBD bioremediation was found to be a potentially profitable, as-yet-untapped production method that would benefit from further research. It is worth noting, however, that the research is limited by its reliance on theoretical models and assumptions, which may not fully reflect real-world conditions, potentially affecting the generalisability of the findings. The lack of empirical validation and practical factors not captured by simulations, such as enzyme stability, further constrain the applicability. Future work should focus on empirical testing and exploring a wider range of parameters to improve the results' relevance and generalisability.
Description
A research report submitted in fulfillment of the requirements for the Master of Science in Engineering, In the Faculty of Engineering and the Built Environment , School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg, 2024
Keywords
UCTD, etrahydrocannabinol, cannabinol, enzyme reactor, bioremediation, SuperPro Designer
Citation
Flavell, Erin Reece . (2024). Design of an Industrial Process for Enzymatic Cannabidiol Conversion [Masters dissertation, University of the Witwatersrand, Johannesburg]. WIReDSpace. https://hdl.handle.net/10539/45383