The total cost of the all the UAEs, depicted by the Batch Generic Box in SuperPro, was the most expensive , followed by the belt press filters . For better cost estimates of the UAEs used in the model, a quote was obtained from an industrial manufacturer named REUS® . Pricing forthis equipment was reduced by about 5%, assuming a conversative approach on the expected discount for bulk purchase of equipment .The OPEX, otherwise referred to as annual operating cost, and COGS are shown with and without depreciation, insurance, and local taxes. The annual operating cost and COGS without including depreciation are more representative of the expected cost for production since depreciation is typically spread over years to effort to expense cost over time while simultaneously lowering the value of the asset . Also depreciation is not a cash outlay so does not have a negative impact on profitability . Here, depreciation was calculated using the straight-line method with a depreciation period of 10 years and salvage value of 5% of the DFC. Insurance was estimated to be 1% of the DFC and local taxes were estimated to be 2% of the DFC. As expected, evaluation of the facility’s annual operating costs and COGS including depreciation, insurance and local taxes are higher compared to without.An analysis was performed to investigate the effect of factors such as depreciation, insurance, factory expenses and local taxes on the economics of the model. A breakdown of the annual operating for each case are shown in Figure 3.4 through 3.6. In all cases, nursery grow bag the annual operating costs are composed of the following: raw materials, labor dependence, facility dependence, Laboratory , Quality Assurance , Quality Control , consumables, waste treatment and utilities.
When the first of the three cases were analyzed, it was determined that the largest contributor to the annual operating costs was the facility dependent cost. Here, facility dependent costs include maintenance, depreciation, insurance, taxes, and factory expenses. Maintenance of each equipment was determined using equipment-specific multipliers, default values provided by SuperPro. No pre-existing depreciation of equipment was assumed in the model. Local taxes were assumed to be 2% of the DFC, in alignment with values from municipal tax charts for South Dakota . Percentages for insurance and factory expenses were estimated using a previous SuperPro model performed internally which collaborated with a plant based bio-manufacturing facility for pricing. The second largest contributor to the annual operating cost was the raw materials. A breakdown of the raw materials and their costs can be seen in Figure 3.7. Here, raw material costs amounted to a value of $5.7 million per year. A list of the raw materials used, the quantity used annually, and their total cost can be seen in Table 3.7. Ethanol, or ethyl alcohol as defined in SuperPro, had the biggest yearly expense out of all the raw materials, totaling over $4.1 million. The model estimates 6.1 million kg of ethanol needed to produce 100 MT of resveratrol. It was determined that the cost of ethanol is over two thirds of the total raw material cost, accounting for 71% of the raw material cost. The price used in the model was $2.00 per gallon of ethanol. This value was retrieved using data by Market Insider which tracks the price of commodities like ethanol daily . For the second case, which only excludes depreciation as a factor contributing to facility dependent costs, the largest contributor was raw materials. Raw materials costs are the same in all three cases but now account for 50.9% of the annual operating cost, with the cost of ethanol remaining the largest contributor.
The second largest raw material cost is the knotweed rhizomes used in the process. Roughly 7.3 million kg of knotweed rhizomes are needed in the process simulation to produce 100 MT of resveratrol. The cost for producing a kg of knotweed rhizomes was estimated to be about $0.19, totaling $1.4 million or 24.3% of the raw materials cost. Estimates for a kg of knotweed were calculated to include costs associated with pre-planting, harvesting, postharvesting, farm equipment operating costs, and cash overhead . A breakdown of these calculations is shown in Chapter 2. Following the second case, which provides a clearer estimate of annual operating cost because depreciation is not included, labor was the third highest contributor to the annual operating cost.Here, labor dependent costs refer to the costs associated with labor hours required to effectively operate the downstream processing section. Labor cost for the upstream production of Japanese knotweed rhizome was already included in the overall unit cost for a kg of knotweed. In this model, the downstream operators earn a wage of $22 per hour. The downstream labor cost includes 40% benefits factor, 10% operating supplies factor, 20% supervision factor, and 60% administration factor. An example of this distribution is as follows: for every $20 paid to an operator for an hour of work, there is an additional cost of $8 for benefits, $2 for supplies, $4 for supervision and $12 for administration. Total labor hours amount to 20,398 hours per year with operators devoting most of their time to the blending tank used as an adsorption vessel. This was in alignment with our estimates since the blending tank is one of the two equipment with the highest number of operations needed compared to every other unit in the model. The next factor contributing to the annual operating cost was the Lab, QA, and QC group, which accounted for less than 1 percent. No funds were allocated to on-going research and development whereas, QA and QC were estimated to be 5% of the total labor cost.
Only one consumable was defined within the process, as listed in Table 3.8. This consumable was the macroporous resin, NKA-11, which is used within the batch adsorption vessel . The annual cost of the resin amounted to $130,536. Pricing for this resin was estimated using commercial values found online by a large scale supplier Co., Ltd.. Waste treatment costs were also incorporated into the annual operating cost calculations. This waste treatment category incorporates the price to safely dispose of different waste streams generated in the facility. Exiting waste streams are classified under one of four groups: organic waste, aqueous waste, solid waste, or gaseous emissions. Here, the cost of emissions was negligible due to low concentration of nitrogen, oxygen, ethanol, and water vapor being released into the atmosphere. The cost to dispose of each group is as follows: organic waste is $0.01/kg, aqueous waste is $0.001/kg and solid waste is $0.01/kg. The annual amount of waste produced by the process is about 44 million kg, estimated at about $234,000 per year in waste treatment costs. While the cost for waste treatment of each group remained similar, the largest contributor to the cost is attributed to organic waste, which include the disposal of the biomass. Resin disposal costs were negligible. These results are in alignment with the cost and use of raw materials throughout the process. A breakdown of the annual waste treatment contributors is shown in Figure 3.8.Utilities are the last factor which contribute to the annual operating cost. A detailed breakdown by utility type used in the process model annually can be seen in Figure 3.9. Three utility types were used in the model: standard power, steam, and chilled water. Pricing for each utility were set as, $0.10 per kW-h of power, $12.00 per MT of steam, and $0.40 per MT of chilled water. The cost of std power was the highest of all three utilities. Std power was used to operate all major equipment including but not limited to the stirred reactor acting as the enzymatichydrolysis unit, plastic growing bag the blending vessel serving as the adsorption vessel, the ultrasonic assisted extractor unit, and the grinder. Surprisingly, the grinder responsible for crushing and homogenizing the plant material into powder required the largest amount of power to operate. Total utilities costs only amounted to a 1% of the annual operating cost. In this model, the annual operating cost for each case described above were used and divided by 100 MT, resulting in a COGS of $150/kg, $111/kg, and $79.4/kg. A summary of both the process parameters and economic analyses from the base case model is shown in Table 3.9.Throughout the design of the simulation model, certain parameters were initialized to match resveratrol production practices described in both patents and scientific literature. When extrapolating bioprocessing parameters from these sources, a conservative approach was taken so that the lower values from a range of data was used in the model in effort to portray realistic results expected during large scale production. An example of this is including a loss of 5-10% material during a filtration step, instead of expecting an 100% recovery every batch. As acknowledged in Chapter 3, certain bioprocessing parameters used in the model were assumptions, thus leading to some uncertainty whether the process outputs were reliable. For this reason, a sensitivity analysis was performed to assess how a certain variation effects the economics of the model, specifically the CAPEX, OPEX, and COGS. Similarly, it is acknowledged that there are multiple designs which can be implemented to produce 100 MT of resveratrol and this model simply serves as an example of one method. Therefore, certain scenario analysis were performed to assess the effect of design and production amount on key economic parameters.To assess the sensitivity of the base case model, we varied certain process parameters to investigate their impact on the CAPEX, OPEX, and ultimately the COGS. One parameter which was defined in the SuperPro model using a conservative approach was the concentration of resveratrol present in the Japanese knotweed rhizome used for processing . Figure 4.1 demonstrates the relationship between COGS and when the concentration of resveratrol per knotweed rhizome is increased up to 3 mg/g FW. The concentration of polydatinwas also increased proportionally along with resveratrol. Each facility simulation is redesigned for each concentration tested while still reaching 100 MT resveratrol annually.As expected, when a larger concentration of resveratrol was modeled to be present within the Japanese knotweed rhizome entering the process, the model resulted in a decrease in both the CAPEX and COGS. The largest drop in CAPEX occurs directly when the concentration of resveratrol is increased an order of magnitude to 1.0 mg/g. The decrease in CAPEX from 0.5 mg/g to 1.0 mg/g is $13 million, over 3-fold larger than the average drop between increments. Expectedly, the COGS also decreases the largest amount between the first two concentrations. The COGS drops a value of $48/kg to a value of $102/kg, a 32% drop. Values for both economic parameters begin to plateau around a concentration of at 1.5 mg resveratrol/g FW, approximately at values of $27.8 million and $88/kg for CAPEX and COGS , respectively. It was discovered that resveratrol is found in a wide range of concentrations in Japanese knotweed rhizomes. A table listing different resveratrol concentrations found in Japanese knotweed is shown in Chapter 2. As mentioned, an average value of the total resveratrol concentration in knotweed, including polydatin, was about 7 folds higher than just free resveratrol. Using information on resveratrol concentrations for Japanese knotweed rhizomes grown specifically in North America1 , we calculated an average concentration value of 2.6 mg resveratrol/g FW. If a future process used rhizomes under similar conditions , our simulation suggests a cost decrease of about one third for CAPEX and 43% for COGS compared to our base case model operating at 0.5 mg Rsv/g. Notably, the same authors that describe an average concentration of 2.6 mg Rsv/g Japanese knotweed roots also mention certain samples contain concentrations as high as 12 mg Rsv/g FW and 12 mg Polydatin/g FW. Using this information, a simulation operating at the same conditions as the base cased was modeled using concentrations of resveratrol and polydatin at a 1:1 ratio at 12mg/g for each stilbene compounds. The same economic analysis was performed on the 12mg/g case, the resulting CAPEX, OPEX, and COGS values are shown in Table 4.1.In the previous chapter, the cost of ethanol was identified to be the major contributor to the annual operating cost and the largest bottleneck.