This study uses USDA Agricultural District 30 (the southwest 14 of the 105 Kansas counties) for the case study. The region is a major contributor to grain and livestock production in Kansas, accounting for 30 percent of the cattle and calves, 26 percent of sorghum, 22 percent of corn, and 15 percent of winter wheat produced in the state. In 2017, it accounted for 8.2% of Kansas farmers (USDA-NASS, 2019a) and 18.2% of Kansas cropland with an average farm size of 347.3 ha (USDA-NASS, 2019b).

The region accounted for 4.7% of the 2019 Kansas population (U.S. Census Bureau, 2019). The region’s population was relatively flat from 2010 through 2015, growing at an average rate of 0.15%, about half of Kansas’s population growth rate of 0.31%. However, the region’s population has been declining at about 0.9% per annum since 2016, compared with an increasing, albeit nearly flat growth rate of 0.03% for Kansas.

The Southwest Kansas agricultural district depends on the Ogallala Aquifer for its irrigation and other water needs. The Ogallala Aquifer’s water levels in the study area are declining (Scott, 2019) though some areas in the western part of the region, such as in Stanton and Morton Counties, reportedly show insignificant changes in water levels. The availability trend and climate change projections suggest the potential intensification of fertilizer use to mitigate the adverse drought effects on crop production in the region (Lindsey, 2013).

The region has good to excellent availability of wind for community-scale energy production (Figure 1). There are few population centers located close to areas with excellent wind resources, reducing the social challenges associated with the development of wind energy in many communities (Gross, 2020). The research focuses on exploiting decreasing wind energy production capital costs, improving ammonia production technologies, and a growing investor confidence in renewable energy economics (Milborrow, 2020), (Sanchez and Martín, 2018; Hauch et al., 2020; Ghiyati, 2021a) to explore opportunities for changing Southwest Kansas’ economic trajectory through local green ammonia production for local use. The results will provide insights for deploying similar solutions in other regions around the world.

The study considers only the cropland allocated to the five principal crops produced in the study area–corn, cotton, soybeans, sorghum, and wheat–and the required anhydrous ammonia (equivalent) used on those crops. The relevant cropland used was the average planted area for the five major crops between 2017 and 2019, inclusive, which accounted for about 64% of the total cropland in the study area (USDA-NASS, 2019b). Ammonia requirements in the study area are assumed to be equivalent to the product of a proportion of the area’s total cropland and the average ammonia requirement per hectare of cropland using 82% nitrogen to convert all ammonia fertilizer products into anhydrous ammonia equivalent. Production and anhydrous ammonia assumptions used for the study area are summarized in Table 1. The total average cropland used is about 1.4 million ha at 0.18 MT/ha of anhydrous ammonia, resulting in an annual requirement of 248,188 MT of anhydrous ammonia for the area. This is assumed as the project’s production target.

Economic feasibility is often assessed using cash flows and specific indicators developed to evaluate the achievement of project objectives. Traditional economic feasibility assessment metrics depending on cash flows include Net Present Value (NPV), Payback Period (PBP), and return on investment (ROI). Given the project’s objective to enhance farm incomes by alleviating producers’ exposure to high anhydrous ammonia prices while contributing to community economic viability, anhydrous ammonia price discount value and ROI for community investors are measured. Relatively competitive thresholds are set for both variables since the project competes with all investment alternatives available to both potential crop producers and community investors in the study area. Further, because of the long duration of the project’s lifespan, all monetary benefits are measured in present value terms to provide a clear performance of the project for potential investors.

NPV recognizes the time value of money by discounting net cash flows from the project at a specified discount rate. The discount rate incorporates potential investors’ perceptions about the project’s risk, and therefore, differs from the interest rate, which is the price of capital. Therefore, the discount rate is generally higher than the interest rate. NPV is formally defined as the sum of discounted cash flow, i.e.,: N P V = ∑ t = 1 T π t 1 + δ t − K 0 (1) where π t and δ define the cash flow in period t and the discount rate, respectively, while K 0 is the capital expenditure assumed to occur in period 0. The project will be able to replace itself at the specified discount rate if the NPV is greater than zero. Closely related to the NPV is the internal rate of return (IRR), which provides the discount rate that produces a zero NPV. The IRR is defined as follows: 0 = ∑ t = 1 T π t 1 + I R R t − K 0 (2)

Some investors have a minimum IRR–hurdle rate (HR)—below which the project is deemed unfavorable for investment. If investors’ hurdle rate exceeds the IRR, then the project requires higher cash flows to be attractive to investors, otherwise, it may not be deemed favorable for investment. Different investors would have different HR for different projects.

It is sensible to expect investors to make a single investment in the project and for the project to not only maintain itself from its cash flows but replace both its energy and ammonia systems at the end of its 25-years lifespan. Another indicator used to specifically measure the extent to which the project can replace itself from its cash flows is the discounted replacement multiplier, RM, estimated as: R M = ∑ t = 1 T = 25 π t 1 + δ − t K 0 (3)

The higher the RM, the more confident investors will be that they would not be asked to make more investments to sustain the project. The discounted payback period, PBP _T , is defined as the earliest time (in years) it takes for cumulative cash flows to equal the initial investment, i.e., P B P T = K 0 ∑ t = 1 T min π t 1 + δ t ≥ 1 (4) where all variables are as defined.

The ROI is an efficiency metric for equity use, comparing the gain from an investment given the amount invested. It measures how well the investments of both producer and community investors do, allowing them to compare their investments in this project with alternative investments available to them. Three specifications of ROI are measured for this project: 1) Producer investors’ ROI; 2) Community investors’ ROI; and 3) Overall project ROI. The overall project ROI is estimated as follows: R O I = ∑ t = 1 T π t 1 + δ − t − K 0 K 0 × 100 % (5) where all variables are as previously defined. The value of the price discount or savings on anhydrous ammonia expenditure as a result of investing in the project is defined for each producer investor i in period t as follows: v i t = α t p t q i t (6) where α t is the price discount rate, p t is the price in period t, and q i t is the quantity of anhydrous ammonia procured by producer investor i in period t. Aggregate producer investors’ return on investment, ROIP, is the value of the price discount they receive throughout the project’s life based on their investment. It is defined as follows: R O I P = ∑ t = 1 T v t 1 + δ − t − K 0 p K 0 p × 100 % (7) where v _t is the value of the price discount accruing to all producer investors in each year and K _0p is the total value of capital producer investors contributed to the project. Aggregate community investors’ return on investment, ROIC, is estimated using their dividend payouts over time, i.e., R O I C = ∑ t = 1 T R t 1 + δ − t − K 0 C K 0 C × 100 % (8) where R t is the dividend payout in each period and K _0C is the total value of capital community investors contributed to the project. The return per share translates the rate measures in Eqs. 6–7 into monetary measures for investors. Producer investor’s value per share (VPS) is the total value of anhydrous ammonia expenditure savings less the producer investment divided by the number of shares, i.e.,: V P S = ∑ t = 1 T v i t 1 + δ − t − K 0 i S i (9) where all variables are as defined, and K 0 i is the capital invested in procuring S _i project shares by investor i. From Equation Eq. 9, the value accruing to each producer investor from their investment would be the product of the VPS and the number of shares they own. Likewise, community investors’ return per share (RPS) is defined as follows: R P S = ∑ t = 1 T R t 1 + δ − t − K 0 i S i (10) where all variables are as defined above. The return accruing to each community investor from their investment would be the product of the RPS and the number of shares they own. The total number of shares is based on the plant’s anhydrous ammonia production capacity, which is 248,188 MT. The total number of shares each investor group owns is determined by its proportional contribution to the project’s initial capital investment. Project funding emanating from debt is distributed to investors on a proportional basis of this initial contribution to capital. Therefore, producer and community investors own equal equity under the 100% Equity Financing condition, 46.67% and 53.33% under the Debt I condition, and 70% and 30% under the Debt II condition.

The sensitivity of project results to the price discount and the dividend rates are explored. The goal is to ensure that investors are getting the highest value from their investments constrained by the project’s ability to replace itself after 25 years. Thus, unlike many farmer/community investments that seek to build capital for the “business”, the investment objective of this project is maintained by paying out returns from the project to investors, retaining only enough to maintain operations.

The study assumes that anhydrous ammonia is produced using locally produced wind energy. The electricity production system uses parameters from NREL’s SAM (System Advisors Model) version 2020.11.29 (Blair et al., 2018) for the Vestas V100–1.8, a 1.8 MW rated turbine manufactured by Vestas Wind Systems (https://www.vestas.com/). Its 50 m blades and 100 m rotor diameter and hub height enable the turbine to deliver a high rotor-to-generator ratio, maximizing productivity in low to medium wind sites. Its three blades sweep an area of 7854.0 m². Figure 2 shows SAM’s simulated average monthly energy output for the V100-1.8 for the study area’s wind parameters. Based on this output profile, the total annual energy output is estimated at 6,997 MWh per annum per turbine. The power curve for the V100-1.8 is presented in Figure 3, showing a cut-in wind speed of about 4.0 m/s and a cut-out wind speed of 20.0 m/s.

The economic analysis in the study considers the conventional Haber-Bosch Process and the SOEC with Exothermal Haber-Bosch Reactor.

Financing the capital is important. Suppose a minimum of 35% of total capital investment (equity) is reserved for local crop producers, then the remainder is assumed financed with community equity investment and/or debt. The debt share of the capital investment is distributed pro rata to the equity shareholders after the debt is paid off. Community investors may be local businesses or individual citizens who embrace the project’s vision of enhancing the communities’ economic viability. Because anhydrous ammonia price risk is a major project objective, it is assumed producer investors receive a price discount and non-producer investors receive dividends, estimated as a percent of net cash flows after any debt payments and anhydrous ammonia price discount. The price discount rate and dividend rate are defined to ensure 1) the project can replace itself after its 25-years useful lifespan (Renewables First. (n.d.)), (Jacobson, 2016), and 2) investors receive returns that adequately justify their participation in the project.

The research explored three financing formulas: 50% producer and 50% community investment (100% Equity Financing); 35% producer and 40% community investment with 25% of debt financing (Debt Financing I); and 35% producer and 15% community investment with 50% of debt financing (Debt Financing II). This fundraising approach was used by Greenfield Nitrogen in Garner, Iowa (Greenfield, 2018), and by numerous producer value-added initiatives, such as ethanol plants.

Ammonia prices are directly influenced by natural gas prices and commodity prices (Schnitkey, 2016), (Ibendahl, 2021). Steadily increasing natural gas prices and volatile corn prices between 1990 and 2020 contributed to anhydrous ammonia prices growing at an average annual rate of approximately 4.5% and exhibiting high volatility (Schnitkey, 2016), (Ag Update, 2018). The weakening of commodity prices in the last decade has contributed to the decline in the ammonia price annual growth rate to 1.4%. However, a reversal in both natural gas and commodity prices in 2021 tripled ammonia prices from about $500 per tonne in August 2021 to $1,600 in February 2022.

The foregoing indicates the challenge of projecting long-term ammonia prices when various forces contribute to its level and volatility. To minimize the risk of projection error, two different anhydrous ammonia price scenarios are investigated for their effect on the economic feasibility of green ammonia production in this study. Each price series projection was derived from the average of 100 randomly generated prices for each of the 25 years of the project’s lifespan using the normal distribution random number generation routine in Microsoft Excel^®. The mean and standard deviation parameters for the two series were, respectively, the average and standard deviation of anhydrous ammonia prices from 2012 to 2020 obtained from the US Geological Survey (US Geological Survey, 2020), and from 2010 to 2020 obtained from the Economic Research Service (Mosheim, 2019). The two projected prices are presented in Figure 5. A time-based analysis of their variability showed that the average standard deviation increased towards the end of the series for both price series. Also, while the figure shows the two price series were different, their correlations coefficient confirmed that they were independent of each other, providing two distinct future price possibilities for consideration. The summary statistics for the two projected price series over the 25-years duration are presented in Table 4. Table 4 shows that the average price for anhydrous ammonia under Scenario 1 was $610.08/MT compared to $475.16/MT for Scenario 2. The correlation coefficients between the two-price series were not statistically significant, suggesting that they represent different expectations about the future of the ammonia market.

The simulated prices generated from the 100 replications above for the 25 years are presented in Figure 5.

Anhydrous ammonia production in the study area was simulated to remain unchanged for the 25-years projections. Total expenditures on anhydrous ammonia over the 25-years project lifespan are approximately $3.79 billion under Scenario 1 and $2.95 billion under Scenario 2. These expenditures are equivalent to annual expenditures of $151.42 million under Scenario 1 and $117.94 million under Scenario 2.

Discounted and undiscounted results under 23% anhydrous ammonia price discount and 30% dividend rate and 100% equity financing of the higher efficiency SOEC technology are presented in this section.

Table 5, presenting the undiscounted results, shows that the project accumulated $1.26 billion at the end of the 25 years under Scenario 1, and nearly $808 million under Scenario 2. Producer investors saved $484 million in anhydrous ammonia expenditure over the 25 years under Price Scenario 1 compared to $292 million under Scenario 2. The dividend payout was structured to produce about the same amount for both groups of investors with 50:50 share ownership. Due to the distribution equity, both investor groups experienced about 124.5% ROI or about $7,000 in return or value per share under Scenario 1. The ROI under Scenario 2 was lower, about 75%. Payback period was estimated at 10 years for Scenario 1 and 12 years for Scenario 2.

Compared to investments that have received significant producer support, such as ethanol plants, this green ammonia production investment seems less risky. This is because its output is a necessary input consumed by its investors, and investors gain control over the price of this necessary input. Indeed, they are promised to pay only a portion of the prevailing market price in each year.

On the critical indicator of replacement multiple, the undiscounted replacement multiplier is above two under both scenarios. Thus, after providing a price discount and paying dividends to investors, the project is still able to accumulate enough cash flow to replace itself twice at the end of 25 years. While undiscounted metrics can be deceiving because a dollar tomorrow is not worth as much as a dollar today, this suggests substantial confidence in the ability to make needed replacements when necessary.

Discounted results under 100% equity financing, presented in Table 6, show that the NPV is positive under all scenarios at a discount rate of 5.0%. The fact that all the indicators meet their threshold conditions suggests that the project is economically feasible under the 100% equity financing situation. Yet, the discounted ROI for investors was about 26% under Scenario 1 and only about 2% under Scenario 2, making the investment uncompetitive with the long-term S&P 500 return of 7% (Maverick, 2020). This means that the investment will be less attractive to investors if Scenario 2 prevails. Because the discounted replacement multiplier is greater than unity under both anhydrous ammonia price scenarios, the project could increase the price discount and dividend rates to increase its attractiveness to potential producers and community investors.

The sensitivity of the returns on investment (ROI) for producer (ROIP) and community (ROIC) investors were investigated with the view of determining the potential effects of increasing investor payouts to increase project attractiveness on overall project economic performance. The limit of increases in these rates is determined by the project’s ability to accrue enough to replace itself, i.e., positive NPV.

The parameters used for the sensitivity analysis were increased by 1 percentage point over 10 steps from their base values of 23% price discount and 30% dividend. The summary statistics are presented in Table 7. While ROIP dominates ROIC for all scenarios after the initial price discount and dividend rates of 23% and 30%, respectively, dividend rate elasticities are higher than price discount elasticities for each scenario (Table 7).

Unlike ROIP and ROIC, increasing the price discount and dividend rates decrease the replacement multiplier. Hence the price and dividend rate elasticities of the replacement multiplier were all negative. For example, the price discount rate elasticity of the replacement multiple was −0.81 under Scenario 1 and −0.83 under Scenario 2. The dividend rate elasticity of the replacement multiple was −1.06 for Scenario 1 and −1.08 for Scenario 2.

The test of the hypothesis that there is no difference between price discount and dividend rate elasticity was rejected for all variables under all scenarios except ROIP under Scenario 2, setting the significance level at 0.05. This provides a strategic direction for project managers on how the attractiveness question may be addressed. Because the dividend rate produces larger absolute elasticities than the price discount rate, it would seem that increasing the dividend rate more than the price discount rate could produce a higher impact on ROIP and ROIC. However, this must be done with a focus on their effect on the replacement multiplier.

Financing the project without debt requires producer and community investors to put up significant capital. Assuming that there are 4,000 eligible “average farmers” in the study area, this would require each of them to put in about $98,000 for 50% equity! At an average anhydrous ammonia price of $500/MT, the average farm must be about 1,062 Ha for the required capital to equal 1 year’s expenditure on anhydrous ammonia.

The difficulties described above for 100% equity financing can be addressed with Debt Financing. Debt Financing I assumes 25% of the project’s capital cost is debt, farmers put up 35%, and the community 40%. The interest rate on debt is assumed at 5.75% and it is financed over 7 years. At this level of farmer equity, the average investment for 4,000 farmers is about $68,174. This was about 38% higher than what the average Iowa farmer invested in ethanol processing facilities in the early 2000s (Institute for Agriculture and Trade Policy, 2006).

The discounted net cash flow from the foregoing financing strategy and other associated performance indicators using a 5% discount rate are presented in Table 8. The 25-years total discounted net cash flow was estimated at $242 million under the first price scenario compared to approximately $24 million under the second price scenario. The discounted total investment value to producers was about $215 million under the first price scenario and about $122 million under the second. The results also indicated that while the conditions under the first price scenario contributed to a replacement multiplier of 1.3, i.e., the turbines and ammonia production plant can be replaced in 25 years from cash flow and investors will have more than 30% of their equity left over after such replacement. On the other hand, replacement under the second price scenario leaves a surplus of only about 3%. Finally, community investor returns on their investment were about four times higher under the first price scenario compared to the second price scenario and about one-and-a-half times under the first price scenario under 100% equity financing.

Debt Financing II tests the sensitivity of the project to debt financing, increasing debt to 50% compared to 25% under Debt Financing I. Producers’ equity contribution remains at 35% but community investment goes down to 15% from 40%.

Increasing the debt used to finance the project reduces the replacement multiplier from about 1.31 under Debt Financing I to 1.15 under the first price scenario and from 1.03 to 0.87 under the second price scenario. Therefore, the project would be unable to replace itself from cash flow if debt financing is doubled to 50% and community financing is reduced from 40% to 15% and the projected ammonia prices under the second price scenario prevail. The project is, however, economically feasible under the first price scenario but not under the second price scenario. Yet, the financing arrangement under Debt Financing II produces a higher return to community investors because the dividend rate is not changed despite the lower contribution from community investors. The foregoing results are summarized in Table 9.

For the project to be economically feasible under both scenarios, the ammonia price discount and dividend rates must be adjusted. Holding the price discount rate at its current rate of 23%, the project can become economically feasible if the dividend rate is set over the 25 years at no higher than 19.6%. That dividend rate drives the discounted net cash flow to zero for Scenario 2 and increases the replacement multiple for Scenario 1 to 1.3 and Scenario 2 to 1.0. If the dividend rate is maintained at 30% and the price discount is adjusted instead, the results show that it has to decrease from 23% to 14.7% for the project to be feasible under both price scenarios. However, because producer return on investment is negative (−7.8%) under Scenario 2 for this condition, it will not be attractive for producers. This means this higher level of debt financing makes the project infeasible.

The foregoing results are for the most efficient (6.5 MWh/MT) energy-to-ammonia SOEC technology system. The results for the lower efficiency energy-to-ammonia SOEC technology (8.1 MWh/MT) are presented in Table 10 under the Debt Financing I conditions, i.e., 25% debt, 35% producer equity, and 40% community investment. The specified energy-to-ammonia rate would require a higher number of turbines to produce the required electricity for the 248,188 MT of ammonia needed annually in the community. The capital intensity under this conversion efficiency was presented in Table 2 above. The table showed that total capital expenditure under the 8.1 MWh/MT efficiency condition was about 25% higher compared to the 6.5 MWh/MT efficiency condition.

Table 10 shows the discounted financial results for the project under the two projected price scenarios. It shows that the project is only feasible under ammonia price scenario 1, posting a discounted net cash flow of about $67.5 million and a replacement multiple of 1.07. While producer investors’ made about 50% return on their investments, community investors only made about 15%. Since the price projection under scenario 1 is fairly optimistic, making investments based on these results carries a higher level of risk than under the more efficient energy-to-ammonia situation. It was determined that for the project to be feasible under both scenarios, the discount rate must be no higher than 3.38%, which, being below the market cost of capital, suggests that the idea is impractical.

Since the 8.1 MWh/MT energy-to-ammonia conversion technology is deemed infeasible for this more advantageous SOEC technology, the conventional technologies are irrelevant alternatives given their higher capital cost outlays to produce the same quantity of anhydrous ammonia under the assumed conditions of prices, interest rate, and discount rate.

This research sought to highlight the potential economic viability that may be engendered by local green ammonia production in small towns and rural (STAR) communities. The local production of green ammonia ensures that money that currently flows out of STAR communities to purchase ammonia will stay within those communities. Local production boosts local economies by creating new jobs, increasing local populations, and fueling economic development. Local green ammonia production also protects farmers from the vicissitudes in the ammonia market, enabling them to have better planning of their production and cost management. Having local production also eliminates supply chain problems that often force farmers to purchase their ammonia well ahead of time without any guarantee of receiving it.

The research focused on returning dividends to investors while providing a guaranteed discount to farmers no matter the market price for ammonia. This ensures that producers in the community have an ammonia fertilizer input cost advantage, allowing them to deal with crop prices a lot more effectively. That the project pays dividends to its community investors implies it improves their cash position, allowing them to support local businesses through reinvestment of their dividends in those businesses or consumption of their products and services. Ultimately, once the replacement cost of the project has been banked, investors and owners may decide to collaborate with their community to recruit businesses that have the potential to enhance the community’s economic viability and also fit its social values. This way, STAR communities have more active control over shaping their future, sustaining or growing their populations, and making themselves attractive as a destination for economic progress in an increasingly competitive market.

Money leaving STAR communities can be invested in local projects. Assuming an equal probability of the two price projections occurring, the projected expenditures on ammonia alone leaving the case community average about $131.2 million per year. With the assumed 25-years lifespan of the project, the community could retain approximately $3.28 billion, holding all other things constant under the assumed conditions of the research. A local anhydrous ammonia plant using local wind energy as its energy source could terminate this capital migration and enhance economic activity in these STAR communities. Additionally, bringing anhydrous ammonia prices under local control decouples its price from traditional natural gas and commodity prices for local producers, thereby reducing the variability that they cause in farm incomes (Huang et al., 2009).