Title: Estimating an Achievable Target Price to Regenerate Bio-Oils Post Hydrogen Sulfide Removal

Bio-oils offer valuable use as bio-solvents for removing hydrogen sulfide (H 2 S) from natural gas. Preceding bench-scale studies indicate that greater than 90% of H 2 S can be removed from a gas stream; economic analysis of such a process is necessary to determine solvent regenerative power required and price limits on a to-be-determined solvent regeneration scheme. With a processing goal 1000 kmol/h of sour gas and removing 99.9% of H 2 S from gas streams at variable feed concentration, design of an absorption unit to process natural gas using bio-oils was carried out through equilibrium stage analysis. Comparison to conventional amine gas treating was used as a cost threshold for gas treatment. The economic viability of using bio-oils as gas sweetening agents depends on capability of regenerating and recycling more than 98% of the soybean oil bio-solvent to compete with amine gas treating, the most popular industrial method.


Introduction
The use of bio-oilsconventional soybean, high oleic soybean, canola, and sunflowerto remove hydrogen sulfide (H2S) from a gaseous mixture has been demonstrated 1 and has potential for use in the natural gas industry to sweeten gas as it is extracted. High concentrations of H2S in natural gas render it sour, with sour gas defined as natural gas containing anywhere from 4 ppmv to thousands of parts per million H2S. 2 Bio-oils have demonstrated the ability to remove up to 90% of gaseous H2S from a gas stream at the bench-scale 1 and could be used in a scheme similar to the industrially prevalent amine gas treating methods. 3,4 In an effort to evaluate if further investigation into bio-oils as gas sweetening agents could yield a viable process, the economic feasibility must be assessed. The present work aims to determine the number of stages required to remove 99.9% of H2S from methane feed gas with varying concentrations of H2S, to examine the capital costs, determine the required regenerative power of the bio-oil solvent, and ultimately determine the maximum cost of regenerating and recycling the bio-oils to be competitive with currently implemented gas processing methods.

Methods
In considering the price of using bio-oils as extraction solvents for treating sour gas and removing H2S, the cost of such a process would need to be less than or equal to the existing industrial processes for treating sour natural gas. Industrial natural gas production plants using amine gas treating methods typically spend $4/GJ on gas treating. 5 To draw a comparison between using biooils such as conventional soybean or high-oleic soybean oil as the gas sweetening solvent, the factors that impact the cost of the process must be verified. The cost of the absorption unitan extraction columnis a key part of the capital cost estimation. By constructing equilibrium stage diagrams an extraction column can be designed, which can then be used to estimate capital costs. Sensitivity analysis, with respect to bio-oil solvent regeneration capability, can also aid in understanding how reliant the process viability is on being able to recycle the bio-oil solvents.
Recovered sulfur could be a lucrative byproduct, as sulfur is valued at more than $200/ton as a chemical building block. 6 As a baseline comparison, the overall cost can be set equal to the current cost of amine gas treatment ($4/GJ); from there, the total maximum cost of bio-oil solvent regeneration/sulfur recovery units can be calculated and the feasibility of developing a process within those cost constraints can be evaluated.
To construct the equilibrium stage diagrams, experimentally determined partition coefficients for H2S in each bio-oil were used. 1 Feed gas concentrations of 40 and 400 ppm H2S in nitrogen were examined with 99.9% removal of H2S as the target. A 40 ppm H2S feed gas concentration would be 2.6 mol% H2S and 97.4% methane; a 400 ppm H2S feed gas concentration would be 20 mol% H2S and 80 mol% methane. These concentrations were chosen to examine the capability of the bio-oil solvents to achieve 99.9% H2S removal from gas with low and high starting H2S concentrations, as well as to compare with bench-scale studies carried out using 40 ppm H2S. A 1000 kmol/h gas feed rate was applied for all simulations, which is on par with flow rates and volumes used industrially in amine gas treating. A graphical solution method 7 was used to determine the optimum flow rate for the bio-oil solvents and the number of equilibrium stages necessary for removing 99.9% of the H2S from different feed gas concentrations using soybean oil or high oleic soybean oil as the sorption solvent. The Kremser method 7,8 was then used to evaluate the percent of H2S absorbed at each theoretical equilibrium stage for different absorbent flow rates and feed gas compositions: the parameters used are shown in Table 1.

Results & Discussion
Economic evaluation of processing sour natural gas using bio-oils is necessary in determining economic viability of such a process before taking efforts to scale up a bench-scale process.
Methods to regenerate and recycle the bio-oils and recover sulfur are not yet fully understood, so a total price for the process is set equal to that of existing technologies (such as amine gas treating).
After subtracting capital costs and solvent costs from the total price limit, the maximum cost of solvent regeneration and sulfur precipitation was calculated. Discussion regarding the feasibility of developing a process to meet these cost constraints is included.
This study aims to investigate the number of stages necessary for 99.9% removal of H2S from varying feed gas concentrations, the price of the extraction column and initial solvent cost, the potential revenue through conversion of H2S and sale of sulfur, and impact of solvent lifetime and regeneration. Figure 1 shows a general schematic of the proposed process. Figure 1. Schematic diagram of a potential process using bio-oil as an extraction solvent to remove H2S from sour natural gas. An extraction column and solvent cost will be the primary initial costs and will be used to set a limit on how much the solvent regeneration/sulfur recovery units could cost in order for the total process to be economically viable.

Estimation of Stages and Stage Efficiency for Recovering H2S as Concentration Varies
To determine the number of stages necessary for handling sour natural gas using bio-oil extraction solvents, varying concentrations of H2S in methane were examined ( Table 1 in Methods). A graphical method was implemented to determine the optimal flow rates of the liquid absorbent feed and the number of stages required to achieve 99.9% removal of H2S from the feed gas.
Partition coefficient values (K) for soybean oil (SBO, K = 0.08) and high oleic soybean oil (HOSBO, K = 0.1) were previously experimentally determined for H2S partitioning in bio-oils. 1 Two concentrations of H2S were chosen for study: 40 ppm (2.6 mol%), a proximate value for the bench scale equilibrium experiments, and 400 ppm (20 mol%), to evaluate capability to sweeten sour gas streams at the more concentrated end of the spectrum. Figure 2 shows the absorption model framework for graphical evaluation. Equilibrium line The graphical solution for (a) 20 mol% H2S absorbed by high oleic soybean oil, (b) 2.6 mol% H2S absorbed by high oleic soybean oil, (c) 20 mol% H2S absorbed by soybean oil, and (d) 2.6 mol% H2S absorbed by soybean oil are shown in Figure 3. Table 2 summarizes the conclusions for 1.5L'min and N depending on the absorbent and feed concentration of H2S.    .

L'min (kmol/h)
Tables 3 and 4 display results for percent of solute absorbed at each stage, N, using 1-15 stages and varied absorbent flow rates, L'. Table 3 is based on using high oleic soybean oil (K = 0.1) as the absorbent while Table 4 shows results for conventional soybean oil (K = 0.08). Table 3. Percent absorption of H2S using high oleic soybean oil as the absorbent. Percent absorption of H2S depends on the absorbent flow rate L' (varied from 101 -250 kmol/h), the partition coefficient (K = 0.1), feed gas flow rate (V = 1000 kmol/h), and is shown for each stage N, up to 15 stages. The values represent the fraction of solute (H2S) absorbed, and ≥99.9% is shaded green, ≥99.0% is shaded yellow, and <99.0% is shaded orange. The target is greater than 99.9% absorption of the H2S. Another way of examining the system is to set a flow rate and compare the amount of H2S absorbed by the different absorbing bio-oils. In this method, the absorbent flow rate (L), feed gas flow rate (V) are fixed and the absorption factor A is dependent on the K value of H2S in each bio-oil. Figure   4 shows the comparison between the bio-oil absorption capacity at different equilibrium stages for two absorbent flow rates, 125 kmol/h and 200 kmol/h. As seen in Figure 4 and in Tables 3 and 4, each oil is capable of attaining 99.9% H2S absorption after some number of stages, N. In the case of the lower flow rate, as shown in Figure 4a, the difference between the oils is larger (root mean square deviation is 0.037) while requiring 14 stages for both oils to achieve > 99.0% H2S sorption. At a higher flow rate (Figure 4b), the oils' H2S absorption more quickly converges (root mean square deviation is 0.018) and only five stages are required to attain > 99.0% sorption of H2S. A balance must be struck between number of stages, optimal flow rate, and the resulting bio-oil solvent use. For the remaining analysis, an absorbent flow rate of 120 kmol/h soybean oil was chosen to limit solvent use while attaining 99.9% removal of H2S in the least number of stages. and find the factor for ratio of vapor hole area to tray active area. Next, Equation 10 was used to find the surface tension factor, FST, which is a function of the surface tension of the liquid, σ.

Column Design and Capital Costs
Having solved equations 7-10, Equation 11 for the flooding capacity parameter, C, can be solved.
In Equation 11, C is the flooding capacity parameter, FST is the surface tension factor solved for in Equation 10, FF is a foaming factor assumed to be 1.0, FHA is the factor for ratio of vapor hole area to tray active area from Equation 9, and CF can be graphically estimated as a function of FLV and plate height, which in this case was assumed to be 24 inches and gave a CF of 0.35. 7,9 Finally, Equation 12 for flooding velocity is solved. Flooding velocity, Uf, is a function of the flooding capacity parameter, C (Equation 11), and the liquid density, ρ L , and the vapor density, ρ V . Table 5 shows the parameters used in Equations 7-12.    Table 5. Ad/A was calculated using Equation 11, and Uf was calculated using Equation 7. The flooding capacity f must be chosen based on desired column performance.
A good rule of thumb is that flooding capacity is often best around 60%, but anywhere from 40-90% is reasonable. 7 In this case, f = 0.8 is used, as that is reasonable for a gas absorption column and will help keep the column size smaller, which keeps costs lower. 7

Solvent Cost
Regeneration of the solvent (soybean or high oleic soybean oil) will be critical to lowering operating costs and developing the present method into a viable process that can compete economically with amine gas treating. The upper acceptable limit for the cost of regeneration will be explored in this discussion. Examining annual solvent cost as a function of the percent of solvent that is regenerated and recycled gives insight into the viability of using bio-oils as extraction solvents. Soybean oil is examined as a case study representative of similar costs and trends with other bio-oils.
Soybean oil availability is dependent on soybeans harvested, and about 10% (by mass) of soybeans harvested is converted into soybean oil, annually, as shown in Figure 5. Soybean oil accounts for 55% of vegetable oil consumption in the United States, with canola oil taking 14% of the market share and other vegetable oils taking less than 10% each. 14 68% of soybean oil is used in food products, 25% in biodiesel and for bioheat, and 7% goes to industrial uses including solvents, paints, plastics and cleaners. 14 The U.S. Soybean Check-off advertises a "Fuel vs. Food: You don't have to choose" message, indicating a national effort to find new industrial and non-food uses of soybean oils. 15 Soybean oil prices are shown in Figure 6.  On average from 2014 -2018, the price of one metric ton of soybean oil was $679, with yearly prices shown in Table 6. Natural gas plants typically operate year-round with 5-10% of forced downtime for maintenance and outages, both scheduled and unscheduled. 16 Figure 7 shows

Adding Value through Sulfur Recovery
In addition to post-absorption recovery of H2S, which would require the bio-oil solvents to be regenerated and recycled, conversion of H2S to elemental sulfur would allow for sulfur to be a secondary product of the process and generate revenue. Sulfur is a valuable chemical building block used to make sulfuric acid and other commodity chemicals and products. The annual demand for sulfur in the United States exceeds 12.7 megatons, and 36% of that is imported annually. 6 The current selling price of sulfur is around $200/ton, and has remained stable above $150/ton since 2014. 6 A relationship can be developed to predict the revenue generated in recovering and selling sulfur.
This relationship is based on the selling price of sulfur ($200/ton), the amount of sulfur recovered out of the potential amount of sulfur recovered (assumed 90%), portion of the year the plant is online (90 -100% of the year, assuming the absorption is running at full capacity for all the time online), and the feed gas concentration of H2S. This relationship is displayed in Figure 8.

Figure 8.
Assuming that 90% of all H2S processed is successfully captured and converted to sulfur, revenues from $1.2 -$10 million annually can be anticipated from sulfur sales. This is dependent on the concentration of H2S in the feed gas being processed throughout the year, as well as the portion of the year the plant is online and operating (assuming the absorption unit is at full capacity and online for the same percent of the year as the plant). Processing more sour gas (400 ppm) would lead to higher revenues than less sour gas (40 ppm). If the plant and absorption unit are online and full capacity 95% of the year, and the average feed gas concentration for the year was 200 ppm H2S, an annual revenue of approximately $5.6 million could be expected. This would correspond to selling 28,000 tons of sulfur per year.
The H2S recovered from the natural gas must be captured and harnessed in some way, as regulations prevent the emission of this hazardous gas to the atmosphere. 2,6 Conversion to sulfur for sale as a chemical building block is a potential lucrative option, and the Claus process for converting gaseous H2S to elemental sulfur is well established. 17,18

Determination of the upper limit for the cost of solvent regeneration
The amount of CH4 available will be higher when the concentration of H2S is lower, as shown in Figure 9. The most recent price for industrial natural gas is $3.54/thousand ft 3 as of February 2020 19 and is used for the revenue calculations shown in Figure 9, assuming that the absorption unit is able to process 1000 kmol/h of natural gas when online, and that 100% of the CH4 in the feed gas is recovered. However, a more sour gas gives less CH4 but more H2S, and overall it is more profitable to process a more sour gas ( Figure 10). Even as the amount of near-pure methane recovered decreases, the increased recovery of sulfur makes up for the loss and in fact is more valuable. Figure 10. Revenue generated by the designed absorption unit operating with 5% downtime annually, 100% recovery of CH4, and 90% recovery and conversion of H2S to elemental sulfur. As feed gas becomes increasingly sour, although amount of sweet gas for sale will decrease, sales of sulfur will increase and increase the total revenue.
Based on 1000 kmol/h gas entering the extraction column, 5% downtime annually (but otherwise at full capacity), an average feed gas concentration of 200 ppm H2S (with 90% sulfur recovery and 100% CH4 recovery), and 100% solvent regenerated, costs and revenues are shown in Table 7. Total profits for the plant will be revenue from selling sulfur and sweet gas minus the fixed costs of capital investments, maintenance and operating costs. While the cost of the extraction column was estimated here, the maximum cost of a regeneration/precipitation unit is to be determined, there may be other costs for heat exchangers, reboilers and condensers, and other installation parts.
Bryan Research & Engineering found that the absorption and stripping columns of amine gas treatment facilities typically comprise less than 45% of the fixed capital costs. 20 Another study of natural gas plants in Canada using amine gas sweetening methods found that absorbers are typically 10% of the fixed equipment cost, a regeneration/stripping column 22%, and other equipment (heat exchanger, reboiler, condenser, misc.) comprise the remainder of the costs. 5 The study also notes that for most natural gas plants, the sweetening process only accounts for 3% of the capital expenses; liquefaction and other operations within the process account for the remainder. Calculated costs for capital expenses are shown in Table 8. Based on a study of Canadian natural gas plants, natural gas is typically processed at a cost of $8/GJwhere $4/GJ is allocated to pipeline/source costs, leaving $4/GJ for actual treatment and liquefaction of the gas to be sold as liquefied natural gas. 5 Based on the simple model in Table 8, which roughly estimates the capital costs and neglects all parts of operating expenditures other than solvent recovery, and assuming the plant has only 5% downtime but otherwise runs at maximum capacity, a continuous cash flow diagram is shown in Figure 11. This model also assumes solvent regeneration and re-use is at 99% and that the cost of processing the gas is even with that of amine gas treating.
After 12 years, the plant would begin to profit. Based on the analysis, over a 20-year plant lifetime, the gas processing cost would be $3.7/GJ, just under the $4/GJ needed to be competitive with amine gas treating and other methods. However, the present model makes several assumptions, including: solvent regeneration and re-use is at 99%, solvent prices do not increase, cost estimates in Table 8 and relationship between cost of different items (i.e. labor = 6x capital costs) are accurate, that natural gas pipelines prices do not increase, that process gas prices do not decrease, that the absorption operation of the plant is running at full capacity all but 5% of each year, and neglects operating costs. A sensitivity analysis showing the relationship between solvent recovery and the cost of processing the gas is shown in Figure 12. Based on the calculations presented, solvent regeneration and re-use would have to exceed 98% to ensure the costs remain low enough for the overall expense of treating the gas to be comparable with amine gas treating. Increased solvent recovery significantly decreases the cost of treating the gas. Figure 12. Sensitivity analysis of the relationship between solvent regeneration and recovery and the cost of processing gas. The cost of amine gas sweetening methods is approximately $4/GJ natural gas. To be competitive, the process designed and presented here would need to achieve greater than 98% solvent regeneration and re-use.
As of yet, a definitive method for stripping the H2S from the soybean oil and bio-solvents and regenerating the bio-oils to be used again is unknown. However, there are a number of methods for recovering H2S gas in aqueous solutions and oils, well-described in the literature. [21][22][23][24] Most methods fall into one of three categories: chemical precipitation, chemical oxidation, or biological oxidation. [21][22][23][24] The key to process viability would be to develop a method that maximizes the percent of the bio-solvent that can be regenerated and re-used. An alternative to the lofty goal of 98% solvent regeneration and recovery is to find other avenues in which costs can be reduced.
While the price of sweetening is well established in Table 8, the cost of other equipment is more loosely correlated and estimated, and a full process simulation including other equipment needs may give better insight into the overall process cost and the solvent regeneration percent that must be obtained.

Conclusion
Soybean oil and high oleic soybean oil were used as case studies to examine viability of using biooils as solvents for extracting H2S from sour natural gas, with an aim of designing an absorption operating that could process 1000 kmol/h of natural gas and remove 99.9% of H2S from feed gas ranging from 40 -400 ppm H2S. Graphical methods and the Kremser method examined the absorption unit and found a trayed tower with 14 stages, a 2 m diameter and 8.5 m height, could successfully meet these goals with a soybean oil absorbent flow rate of 120 kmol/h. The cost of such an extraction column was estimated along with other capital costs, and the dependence on the economic viability of such process hinges on capability of regenerating and recycling more than 98% of the soybean oil bio-solvent to stay competitive with amine gas treating, the most popular industrial method. The potential of the proposed process being economically favorable looks unlikely. However, we show here where cost-cutting measures could be developed and implemented to reduce processing cost so that bio-based oils are able to compete with current technologies.