The 6 Percent Problem: What 113 Stalled USDA Fertilizer Projects Teach You About the Feasibility Study FIELDS Now Requires

Between 2023 and 2025, the U.S. Department of Agriculture put $517 million into 76 fertilizer production facilities across 34 states and Puerto Rico. The projected result was 11.8 million additional tons of annual fertilizer production and more than 1,300 jobs [1].

On July 1, 2026, Secretary Brooke Rollins stood up and said the previous administration had managed "only building 6% of their stated goal" [2]. The fuller version, relayed through trade press from her remarks: of 121 fertilizer projects identified under the Fertilizer Production Expansion Program, eight were completed. Ninety of them never had a signed agreement at all. Seven were flagged unsatisfactory, together worth $70 million, and in many cases USDA had no contact information on file for the awardee [3][4].

Eight divided by 121 is 6.6 percent. The arithmetic checks out.

That number is now the organizing fact of federal fertilizer policy. The $500 million FIELDS Program, launched the same day, was built directly on USDA's post-mortem of those 121 projects [2]. And the single largest change between the old program and the new one is that FIELDS requires a feasibility study from every applicant, at every dollar level, and reserves the right to reject any application whose study the agency does not find adequate.

This article is the forensics. What the 121 projects actually were. Why most of them never turned into a plant. What FIELDS scores and how many points each thing is worth. The eligibility tables USDA requires you to certify against but has not published anywhere. And what a USDA feasibility study has to contain to survive contact with a Rural Development reviewer.

If you are scoping a fertilizer project, this is the map. If you are a lender or a credit officer looking at one, this is the diligence list.

Part 1: What the money actually bought

FPEP was authorized under Section 5 of the Commodity Credit Corporation Charter Act. USDA opened applications in September 2022 with $250 million, doubled it to $500 million, then raised the ceiling toward $900 million in June 2023. Demand was enormous. Secretary Vilsack reported $3 billion in applications from more than 350 independent businesses across 47 states and two territories for the first two rounds alone [5].

The program's design choices are what matter in hindsight:

• Maximum award $100 million, minimum $1 million

• Five-year grant term

• No matching funds requirement

• A feasibility study required only if you asked for $5 million or more [6]

• Grant funds paid out largely on completion or reimbursement, and critically, developers could not draw money until all of their other financing was already secured

Read those last two together and you can predict the outcome. USDA handed out grant awards to sponsors who had not been required to demonstrate they could fund the other half, then told them the cash would not arrive until they had funded the other half.

The award roster shows the split clearly. Rounds went out across FY2023, FY2024 and FY2025 with cumulative announcements climbing from roughly $29 million and 8 projects in March 2023, to $121 million and 33 projects by October 2023, to $286.6 million and 64 projects by August 2024, and finally $517 million across 76 facilities by December 2024 [1].

The recipients fall into three obvious tiers.

Tier one: incumbent operators doing bolt-ons. AdvanSix took $11.8 million for a 195,000 ton per year ammonium sulfate expansion at Hopewell, Virginia. Pursell Agri-Tech took $5 million for controlled-release granulation. Sul4r-Plus took $14.9 million and actually opened a 100,000 ton per year synthetic gypsum upcycling plant at Marissa, Illinois in April 2025 with 25 jobs. Ostara took $7.6 million for struvite recovery in St. Louis and is operating. These are the ones that worked, and they worked because the sponsor already had a balance sheet, a plant, a workforce and a customer list.

Tier two: large speculative greenfields. Atlas Agro took $80 million for a 700,000 ton per year green ammonia plant at Richland, Washington that needs roughly 320 MW of power and was tied to a hydrogen hub that later lost its funding. Michigan Potash took $80 million for a solution-mining project whose federal environmental review is estimated to complete on October 2, 2026, nearly two years after the award. LSB Industries was offered $77 million and saw it cut to $23 million after it dropped the new urea plant from the scope [7]. Greenfield Nitrogen took $40 million for a low-carbon ammonia plant in Iowa and got stuck.

Tier three: small, novel, and in several cases gone. Vermicompost, black soldier fly frass, algae, enzymatic food-waste digestion, chitosan, humic acid, kombucha-adjacent soil biologicals. California Safe Soil took $3.98 million and ceased operations in May 2025. Rollins mocked this tier publicly, referring to projects involving "worm worms and flower pots and kombucha" [8].

Here is the honest caveat, and it matters for anyone citing these numbers: for the large majority of the 76 facilities, there is no dated public record of what happened. These are small private companies with no SEC filings, no press office and, per USDA's own account, sometimes no working contact details. The same opacity that let projects disappear makes independent verification impossible. What can be said with confidence is what USDA itself has said: eight completed, ninety without signed agreements.

A note on the two numbers. USDA's official investment figure is $517 million across 76 facilities. Rollins's failure figure uses 121 projects. Those are different universes. The 121 counts projects identified or offered awards; the 76 counts facilities with finalized investment. Some coverage has cited roughly $800 million against 121 projects, which conflates the authorized ceiling with obligated dollars [9]. If you quote this story, quote it precisely: eight of 121 identified projects completed, and separately, $517 million obligated across 76 finalized facilities. Do not present "6 percent" as a tons-of-capacity figure. The public record does not support that reading.

Part 2: The seven ways a funded fertilizer project dies

Strip the anecdotes down and the failure modes are structural, not moral. Ranked by how often they appear in the documented cases:

1. The match was never assembled. This is the binding constraint and it is not close. FPEP required no match, so awards flowed to sponsors who had never proven they could raise the rest. Then the drawdown rule made the grant unusable until they did. Linda Thrasher of Greenfield Nitrogen described it exactly: developers cannot access grant money until they have all their other financing secured, so "it is the proverbial chicken or egg problem. The only way you get to the finish line is if you have someone financing who will fund the whole thing" [7].

2. Off-take was never secured. A fertilizer plant without a distribution channel is a pile of steel. Thrasher reported that not a single cooperative would participate in off-take arrangements, and that some told their farmers not to, out of concern about retribution from incumbent suppliers [7]. That is a market-structure barrier a grant cannot solve.

3. Power and interconnection. Atlas Agro's 320 MW requirement is the clean example. Electrochemical and low-carbon routes are exposed to a grid queue that runs on its own timetable.

4. Permitting and environmental review. Michigan Potash's review completing in October 2026 against a 2025 award is not scandal, it is arithmetic. If your entitlement path is not mapped before the application, your construction timeline is fiction.

5. Post-award scope cuts. LSB's $77 million to $23 million reduction shows what happens when project economics move between application and final investment decision.

6. Sponsor insolvency. California Safe Soil raised $19.24 million lifetime and still ceased operations.

7. Technology not proven at commercial scale. Concentrated almost entirely in tier three.

Now hold that list next to what a feasibility study is supposed to evidence. Signed feedstock supply. Signed or letter-of-intent off-take. Committed match with proof of funds. Contractor pricing rather than desktop capital estimates. A permit pathway with a named agency and a real timeline. Sensitivity to a nutrient price collapse.

FPEP's failures were, almost without exception, failures to evidence exactly those six things. That is not a coincidence, and USDA clearly reached the same conclusion.

There is a counter-argument worth stating fairly. Most FPEP awards landed late in 2024, grants were frozen in early 2025, and the five-year grant term means many projects measured as incomplete in mid-2026 are still inside a normal construction window. The Fertilizer Institute worked with USDA to identify recipients with unused dollars and the best potential to deliver, which implies a redirectable pipeline rather than a dead one. None of that rebuts the completion count. It reframes it as a program design failure rather than proof the projects were frivolous.

Part 3: What FIELDS changed

The Fertilizer Investment and Expansion for Long-term Domestic Supply Program was announced July 1, 2026. It carries $500 million in Commodity Credit Corporation funds, administered by Rural Development's Rural Business-Cooperative Service under notice RD-RBS-26-01-FIELDS [2].

The headline terms:

Term	FPEP (predecessor)	FIELDS
Total funding	~$517M obligated	$500M
Award range	$1M to $100M	$15M to $150M
Matching funds	None required	50% non-federal, mandatory
Feasibility study	Required at $5M+ requests	Required for every application
Financing test	Drawdown blocked until financing closed	Match must be secured before the award agreement is executed
Anticipated awards	76 facilities	Approximately 10
Period of performance	5 years	60 months, plus up to 24 months no-cost extension

Three of those changes do almost all the work.

The 50 percent match converts the chicken-and-egg trap into a front-end screen. Under FPEP the financing test happened at drawdown, after the award, when it was too late to reallocate the money. Under FIELDS, unsecured match makes the project ineligible at application. USDA is no longer discovering financing failure eighteen months after announcing a grant.

The award floor of $15 million is a scale filter. At a 50 percent match, the smallest project FIELDS will touch is roughly $30 million in eligible costs. That deliberately excludes tier three.

The universal feasibility study requirement removes the $5 million loophole. Every applicant now has to put a third-party document in front of a reviewer.

Two live discrepancies you need to get right. First, the deadline. USDA's July 1 press release says applications are due August 15, 2026. The notice itself, the Rural Development program page and the Grants.gov posting all say 11:59 p.m. Eastern on August 17, 2026 [2][10]. The August 15 figure appears to have been a press-release error from the start. Plan to the earlier date and submit against the later one.

Second, the ceiling. The notice PDF hosted on rd.usda.gov still reads "up to $100 million." The Grants.gov posting, updated July 7, 2026, raised it to $150 million, added the restriction that pre-development costs cannot be the primary use of funds, and removed the packaging and labeling eligible-use language [10]. The two USDA-hosted files are out of sync. Treat the Grants.gov "Updated 7.7.26" file as controlling. If you sized your request off the rd.usda.gov PDF, resize it.

Part 4: The scoring rubric, decoded

This is where most applicants guess. They should not have to.

The notice assigns 66 base points across four scored criteria, plus up to 20 more in bonus and discretionary points [10]:

Criterion	Points	What earns them
1. Financial Viability and Technical Merit	0-25	Financial projections with supported assumptions; technical merit and technology track record; financing and capital secured or identified. The notice is explicit that applicants with confirmed commitments for matching funds receive more points. Scoring bands run 0 (not addressed), 1-8 (inadequate), 9-17 (adequate), 18-25 (thorough, high likelihood of success).
2. Market Impact and Market Demand	0-30	Split into Market Impact 0-12 (input supply plan, capacity expansion, competition effects, producers and acres benefited) and Market Demand and Opportunities 0-18 (validated demand, production availability, market drivers, partnerships, end-user commitments, project sustainability).
3. Work Plan and Budget	0-25 header, 19 enumerated	Work Plan 0-7 (activities, sequencing, milestones), Budget 0-6 (line-item detail, reasonable and allocable), Key Personnel and Service Providers 0-6 (experience, qualifications, capacity).
4. Major Risks and Mitigation	0-6	Construction delay, equipment availability, labor shortage, service provider performance, operating capital, market loss, legal and regulatory, zoning, environmental, waste and utility risk.
5. Nitrogen or Sulfur Products	up to 5 bonus	Projects process-manufacturing nitrogen or sulfur related fertilizer.
6. Executed Off-Take Agreements	up to 5 bonus	Executed agreements weighted above agreements under negotiation.
7. Administrator Points	0-10 discretionary	Rural economic recovery; applicants that have not previously received an FPEP award; geographic diversity.

Two observations that matter operationally.

The rubric has an internal inconsistency. Criterion 3 carries a header of "0-25 points" but its enumerated sub-parts sum to 19. Some published reconstructions quietly force the total to a round 100. The notice does not. Build to the enumerated sub-parts and do not assume the extra six points exist.

Criterion 7 explicitly rewards not having taken FPEP money. If you are a prior FPEP awardee, you can still apply, but only for a demonstrably separate and additive project, and you are giving up ground on up to ten discretionary points. That is the 6 percent problem showing up directly in the scoring sheet.

Separately, a set of pass or fail gates sit outside the rubric entirely: domestic ownership, independent ownership and operation, the market share test covered below, active SAM registration, no debarment or federal debt delinquency, secured match, and a complete application. The notice states plainly that failure to address any application criterion by the deadline results in the application being determined ineligible.

Part 5: What the feasibility study can and cannot carry

Here is the map from points to paper. It is the single most useful table in this article for anyone actually assembling an application.

Point category	What earns the points	Can the feasibility study carry it?
Financial projections (Crit 1)	Two years pro forma with assumptions, three years historical statements, current statement within 90 days	Yes. The financial feasibility section is the core
Technical merit (Crit 1)	Engineering report, process description, technology track record	Yes. Technical feasibility section, strengthened by a separate engineering report
Financing secured (Crit 1)	Proof of funds, executed term sheets, equity commitment letters	No. The study can describe the stack; points require third-party commitment documents
Market Impact (Crit 2)	Input supply plan, competition analysis, producers and acres served, support letters	Partly. Analysis yes, producer commitments no
Market Demand (Crit 2)	Validated demand data, end-user commitments, distributor arrangements	Analysis yes, commitments no
Work Plan (Crit 3)	Project schedule, milestones, sequencing	Partly. Contractor schedule is external
Budget (Crit 3)	Line-item budget, contractor bid or GMP contract, equipment quotes	No. Applicant and contractor documents
Key Personnel (Crit 3)	Resumes, construction labor, operational workforce	Yes. Management feasibility section plus required resumes
Major Risks (Crit 4)	Risk register with mitigation, permit receipts, sensitivity analysis	Yes. Sensitivity analysis and risk discussion
Off-take bonus (Crit 6)	Executed off-take agreement	No. External contract

Add it up. The two largest scored blocks, financial viability at 25 and market at 30, are 55 of the 66 base points, and a rigorous feasibility study does most of the work in both. But the highest-value sub-criteria inside them, financing secured and executed off-take, require documents the study cannot manufacture.

A feasibility study is necessary and not sufficient. Anyone selling you one as a complete application is selling you a problem.

Part 6: The eligibility gate USDA never published

This is the part nobody else has done, and it is the reason this article exists.

The FIELDS notice requires that applicants, including their affiliates, "must not hold a market share in production greater than or equal to the entity that holds the fourth largest share of that market" for nitrogen, sulfur, phosphate, potash, or any combination [10]. Market share is defined as production capacity divided by total industry production or distribution capacity, times 100. Both a global and a domestic share must be calculated and certified. Affiliation follows the Small Business Administration rule at 13 CFR 121.103.

Every applicant has to certify against a threshold. USDA does not publish the reference tables. So here they are, built from company filings, USGS data and named industry sources.

Nitrogen (U.S. ammonia basis)

Rank	Company	U.S. ammonia capacity	Approx. share
1	CF Industries	~10.4 to 10.5M tons North America; produced 9.8M tons gross ammonia in 2024	~40% of NA capacity [11]
2	Koch Ag and Energy Solutions	~3.5 to 4.5M tons estimated, five U.S. plants	not cleanly published
3	Nutrien, U.S. operations	~3.5 to 4M tonnes, U.S. subset of ~7.3M global	not cleanly published
4	LSB Industries	~900,000 tons/yr	~4.7% of U.S. capacity [12]

LSB's per-plant nameplate from its own Form 10-K: El Dorado, Arkansas at roughly 493,000 tons per year, Pryor, Oklahoma at 246,000, Cherokee, Alabama at 188,000. That sums to about 927,000 tons. Against a total U.S. ammonia capacity of roughly 19.2 million tons, LSB sits near 4.7 percent.

The practical threshold for a nitrogen applicant: stay below roughly 4.7 percent of U.S. ammonia capacity, which in plain terms means do not be LSB-sized or larger.

One honest caveat. CVR Partners, running Coffeyville, Kansas and East Dubuque, Illinois, may exceed LSB on a raw ammonia basis, which would push LSB to fifth. LSB's own investor materials describe the company as fifth largest in North America. The fourth-place threshold is approximately, not exactly, LSB's 900,000 tons. For a new entrant the ambiguity is irrelevant, because both figures sit far above anything a FIELDS applicant would build. Note also that Waggaman, Louisiana at 880,000 tons per year is now CF's following its acquisition from Dyno Nobel and is no longer a standalone producer.

Do not use USGS ammonia figures as your denominator without adjusting them. USGS reports nitrogen content and production, not product-ton capacity, and puts U.S. output near 14,000 thousand metric tons of nitrogen content in 2025 from 18 companies at 38 plants across 19 states, running at roughly 80 percent of rated capacity [13]. That utilization figure is itself worth remembering when you build a ramp assumption.

Phosphate (U.S., phosphoric acid basis)

Rank	Company	Position
1	Mosaic	~60% of North American phosphoric acid capacity, ~7% of world [14]; currently states ~74% of North American concentrated phosphate production
2	Nutrien	Aurora NC, White Springs FL
3	Simplot	Idaho and Wyoming integrated
4	Itafos	far below the top three

Industry analysis has long described the U.S. structure as a "big three" of Mosaic, Nutrien and Simplot plus a disruptive entrant in Itafos [15]. U.S. phosphate rock production runs around 20 million short tons annually, heavily concentrated in Florida. The fourth-place threshold here is a small fraction of North American capacity and is effectively non-binding for any new entrant.

Potash

Global nameplate is roughly 103 million tonnes. Nutrien holds about 20 percent, Uralkali and Belaruskali about 14 percent each, with Mosaic rounding out the top group alongside K+S and ICL. Domestically the picture is trivial: two companies in New Mexico plus Utah operations, and the United States imports roughly 95 to 98 percent of its potash, about 85 percent of it from Canada [16]. Non-binding for any new domestic producer.

Sulfur

U.S. production was 8.20 million metric tons of sulfur content in 2024, almost entirely recovered byproduct from refineries and gas plants. The historic leaders are oil majors, with the top seven accounting for roughly 72 percent of recovered sulfur. The fourth-largest sulfur producer is a refiner at Marathon Petroleum scale. A fertilizer-sulfur processor will not come close. Worth flagging conceptually: sulfur capacity is really refinery throughput, not a fertilizer decision variable at all.

What this means

The eligibility gate is real, and for the independent projects FIELDS is aimed at, it is a formality. Document it, certify it, move on. It binds only if your affiliate group already owns LSB-scale nitrogen capacity. If your segment is distorted, for example by two dominant firms making third and fourth place misleading, the notice provides a justification pathway, and Capper-Volstead farmer associations have their own. Use them early rather than at submission.

Part 7: What a USDA feasibility study actually is

This is where the article stops being about fertilizer and starts being about every USDA Rural Development capital program, because the definition is shared.

The controlling language

FIELDS requires "a recently completed feasibility study, prepared for the project and dated no more than 3 years before submission, signed by a qualified consultant," which must "demonstrate that the proposed project or operation is practical and viable and assess its likelihood of success." And then the sentence that decides applications: "The Agency must concur that the feasibility study is acceptable and adequate" [10].

That makes the study both a pass or fail eligibility item and a scored input. If the agency does not concur, the application is not responsive and will not be considered for funding.

Who counts as a qualified independent consultant

The notice defines a qualified consultant as "an independent third-party person possessing the knowledge, expertise, and experience to perform the specific task required" [10]. The regulation at 7 CFR 5001.3 uses the same construction: a feasibility study is a report including an opinion or finding, conducted by an independent qualified consultant, evaluating the economic, market, technical, financial and management feasibility of the proposed project [17].

Three things follow that applicants routinely get wrong:

There is no approved consultant list. Not in the notice, not in the guide, not in the regulation. Anyone advertising USDA pre-approval is advertising something that does not exist. "Qualified" is judged by the reviewer against your specific project.

Independence is a real test. The consultant cannot hold a financial interest in the project, the applicant or the outcome. Your engineer of record, your equity partner, and your in-house finance team are not independent third parties for this purpose.

Qualifications must be documented in the study. The feasibility study guide requires a resume or statement of qualifications for the author, including prior experience, in the references section. A study without it is incomplete on its face.

The five dimensions

Every USDA feasibility study, across programs, is built on the same five-part framework from 7 CFR 5001.3 and Appendix A to Subpart D:

1. Economic feasibility. Cost-benefit analysis, and the project's effect on suppliers, customers, producers and the supply chain, including its resistance to shocks.

2. Market feasibility. Current and future market potential, competition, sales estimates, buyers and users, market share.

3. Technical feasibility. Reliability of the technology, commercial availability, transportation, utilities, labor, construction risk.

4. Financial feasibility. Ability to generate sufficient income and cash flow to sustain operations long term and service debt.

5. Management feasibility. Legal structure, ownership, board, management analysis, qualifications.

The FIELDS feasibility study guide, published June 30, 2026, wraps those five in an executive summary and a consultant recommendation, plus a references section, for seven elements total.

The change from FPEP is instructive. The FPEP guide used the identical seven-element architecture but only triggered at a $5 million grant request. FIELDS keeps the architecture and removes the trigger. Everyone needs one now.

What the agency does with it

Rural Business-Cooperative Service reviews for completeness and eligibility, then scores. The notice commits to written acknowledgment of receipt within 30 days and anticipates awards in the December 2026 to January 2027 window, though officials reportedly floated September or October verbally at launch. Only the written window is controlling. An adverse decision is appealable to the National Appeals Division under 7 CFR part 11.

Critically, the agency may request clarifying information but will not solicit or consider new scoring or eligibility information submitted after the deadline. A study that arrives thin cannot be topped up later.

Does USDA actually reject studies?

The honest answer is that the public record is thin, and anyone telling you otherwise is likely inventing precedent. National Appeals Division determinations are redacted and searchable only through the NAD portal, not indexed publicly, and no federal court opinion on a Rural Development business-program feasibility dispute surfaced in this research.

What does exist is a 1999 Government Accountability Office review of Rural Business-Cooperative Service loan losses. Of 24 sampled loss borrowers, the agency had not followed its own requirements for 18. Eleven loans were missing feasibility studies entirely, three had studies with significant flaws, and four more were questionable. GAO recommended USDA clarify when it expects feasibility studies and emphasize carefully evaluating them before lending [18]. USDA responded with an administrative notice in January 2000 and nationwide field-staff training on how to distinguish a properly completed feasibility study from a business plan.

That last phrase is the enduring lesson. The most common defect is not a bad study, it is a business plan wearing a study's cover page. A business plan argues the sponsor's case. A feasibility study renders an independent opinion, including the opinion that a project may not work.

Consistent with that tightening posture, the share of Business and Industry applications approved fell from 89.1 percent in FY2021 to 52.7 percent in FY2023, the most recent published throughput data [19].

Part 8: Does the plant actually pencil?

A feasibility study that concludes "yes" without doing this arithmetic is the kind that produced 113 stalled projects.

Capital cost per annual ton

Named precedents, which is what a reviewer wants rather than rules of thumb:

• Iowa Fertilizer Company, Wever, Iowa. Greenfield nitrogen complex, widely reported at roughly $3 billion against an initial $1.4 billion estimate, producing about 2 million metric tons per year of finished nitrogen products including 850,000 tons of ammonia and 460,000 tons of granular urea. Koch acquired it for $3.6 billion. Implied capital cost lands near $1,500 to $1,900 per annual ton of product, considerably higher on an ammonia-ton basis.

• CF Industries Donaldsonville and Port Neal expansions. Board-authorized at $3.8 billion, escalating past $4.2 billion, for 2.1 million tons per year of new gross ammonia plus upgrading. Roughly $2,000 per annual ton before escalation. These figures come from SEC filings rather than press releases, which makes them the most reliable anchor available.

• CF Blue Point, Louisiana. Approximately $4 billion for 1.4 million metric tons per year of low-carbon ammonia, starting 2029. About $2,850 per annual ton, though this is a forward-looking sponsor estimate, not a completed cost.

• Green ammonia premium. A 1,000 tonne per day grey ammonia plant runs roughly $500 to $700 million; a comparable green facility runs $900 million to $1.2 billion. A 40 to 100 percent premium.

Rough order-of-magnitude ladder for 2026, with the caveat that the smaller routes rely on trade and vendor sourcing rather than audited project costs:

Route	Approx. $/annual ton of capacity
Dry bulk blending	$100 to $300
Liquid manufacture and storage	$200 to $500
Granulation, NPK	$400 to $900
Manure, litter and digestate drying and pelletizing	$500 to $1,500
Sulfuric acid	$400 to $900
Ammonium sulfate	$600 to $1,200
Urea, as an add-on to ammonia	$800 to $1,500
Nitric acid and ammonium nitrate	$800 to $1,500
Specialty and controlled-release coating	$1,000 to $2,500
Anaerobic digestion with nutrient recovery	$1,000 to $3,000
Ammonia, world scale	$2,000 to $3,000
Phosphate, integrated rock to DAP or MAP	$2,500 to $4,500
Ammonia, small or modular	$3,500 to $6,000+
Potash, greenfield	above $2,300 per tonne

Escalate every pre-2020 precedent before you use it. The Chemical Engineering Plant Cost Index rose roughly 34 percent from 2020 to 2023, with further increases through 2025 and into 2026. A 2018 capital figure applied to a 2026 project without escalation is not conservative, it is wrong.

Operating cost

For nitrogen the answer is gas and not much else. A modern steam methane reforming plant consumes roughly 30.4 MMBtu per tonne of ammonia. CF Industries realized an average natural gas cost of $3.31 per MMBtu in 2025, up from $2.40 in 2024, and EIA expects Henry Hub near $3.60 across 2026 and 2027. Feedstock alone therefore runs roughly $100 to $110 per ton of ammonia before fuel gas, losses, and the spread between Henry Hub and delivered industrial gas at your actual site.

Scale is the other half. CF runs its entire 10.4 million ton network with 2,800 employees. A sub-scale independent carries the same categories of fixed cost across a fraction of the tons. That structural disadvantage is the reason the incumbents are incumbents, and no grant changes it.

Revenue, and the trap in 2026 pricing

National retail averages, in dollars per ton:

2026	Urea	Anhydrous	UAN32	DAP	Potash
Mid-March	674	924	489	851	488
Mid-May	864	1,126	597	913	493
Late July	682	967	465	911	494

Source: DTN retail fertilizer index [20].

Wholesale told the same story more violently. Urea moved from roughly $400 per metric ton to above $850 in April 2026, then back to $453 by June. DAP climbed from about $580 to $770 [21].

This is a war premium, not a market. Strikes beginning late February 2026 effectively closed the Strait of Hormuz, through which roughly a third of global seaborne fertilizer trade moves. QatarEnergy halted urea production after its Ras Laffan operations were struck, and Qatar's QAFCO alone represents around 14 percent of global urea trade [22]. CF Industries' own filings note Middle East nitrogen at roughly 35 to 40 percent of globally traded urea and 25 to 30 percent of traded ammonia.

An applicant who builds a pro forma off May 2026 prices is doing the same thing the 2022-vintage FPEP applicants did. Base case should be mid-cycle: urea NOLA in the $300 to $350 per tonne range, anhydrous retail $600 to $700 per ton, gas at the forward strip near $3.60 to $4.00, and an 80 percent utilization ramp consistent with how the U.S. fleet actually runs.

The minimum viable plant

Now the arithmetic that decides whether FIELDS is even the right instrument.

A $15 million minimum award at a mandatory 50 percent match implies a project of at least $30 million in eligible costs. The $150 million ceiling implies roughly $300 million.

Routes that fit the floor at $30 to $60 million: dry blending, liquid manufacture and storage, granulation, specialty and controlled-release, ammonium sulfate finishing, manure and litter and digestate drying and pelletizing, struvite recovery, small anaerobic digestion with nutrient recovery, biochar and pyrolysis, torrefaction, and certified organic lines. At $100 to $2,500 per annual ton these reach real tonnage inside the envelope.

Routes that cannot use even the ceiling as a basis: conventional world-scale ammonia and urea at $2 to $4 billion, integrated phosphate at $1 billion and up, and conventional or solution potash in the multi-billions. A $150 million grant covers under 10 percent of a $2 billion ammonia complex. For those projects FIELDS is a top-up inside a much larger independently financed stack, not the reason the project happens.

Marginal: small and modular ammonia, nitric acid and ammonium nitrate, sulfuric acid, and green ammonia pilots may fit the upper award range at genuinely small nameplate, but at $3,500 to $6,000 per annual ton the unit economics are thin.

The credit test underneath all of it: realized netback minus operating cost per ton equals gross margin per ton, times tons per year equals EBITDA, divided by annual debt service equals coverage. Project finance wants above 1.3 times. Run it at mid-cycle prices, not at the spike, and run it at 80 percent utilization, not nameplate.

Part 9: The policy layer, and how much of it to believe

Four things happened in 2026 that a feasibility study has to address, and they are not equally durable.

Critical minerals designation. Phosphate and potash were added to the final 2025 Critical Minerals List on November 7, 2025 [23]. This is structural and it matters, unlocking permitting priority and eligibility for federal support. Michigan Potash's conditional Department of Energy loan guarantee above $1 billion is the visible example.

The Moroccan phosphate duty suspension. The President declared a fertilizer supply emergency and suspended countervailing duties on Moroccan phosphate for eight months in late June 2026. The duty history runs 19.97 percent in 2021, cut to 2.12 percent in 2023, raised to 16.81 percent in 2024. A Texas A&M Agricultural and Food Policy Center study estimated the duties added roughly $6.9 billion to U.S. field-crop input costs between 2021 and 2025, and USDA projected the suspension could cut prices by around 22 percent [24]. But this is temporary and contested. Commerce's June 30, 2026 sunset review recommended keeping the duty, with final results expected around late October 2026. Do not build a permanent phosphate price assumption on an eight-month suspension that the administering agency has recommended against continuing.

Antitrust. A USDA and Department of Justice memorandum of understanding on competition in agricultural inputs was signed September 26, 2025, covering feed, fertilizer, fuel, seed and equipment [25], with an FTC fertilizer investigation reported in May 2026. Implemented, enforcement pending, effect unknown.

The Hormuz shock. Market event, not policy, and already reversing.

Then the demand side, which is the part sponsors most often skip. USDA's Economic Research Service forecasts crop cash receipts at $240.8 billion in 2026, up 1.2 percent nominally but a decline in real terms [26]. Farm-level returns are negative across much of the Corn Belt. Fertilizer runs roughly 40 percent of corn operating cost, with Illinois projecting around $229 per acre on corn for 2026. Grain-to-fertilizer exchange ratios have sat near multi-year highs since 2022.

Your customer is under margin pressure and is actively trimming rates. A demand forecast that assumes volume growth needs to say why.

Part 10: The wider USDA feasibility study landscape

FIELDS is one door. If you are evaluating a capital project in rural America, here is the full stack and where a feasibility study sits in each.

Program	Study required?	Trigger	Program funds the study?	Size	Match or equity
FIELDS	Yes, all applications	No floor	No	$15M to $150M	50% non-federal
B&I Guaranteed Loans (7 CFR 5001)	Yes for guarantees above $1M to a new entity or new activity; agency discretion otherwise	$1,000,000	No	To $25M per borrower, $40M for value-added ag cooperatives at Secretary discretion	Graduated 10% to 25% balance-sheet equity
REAP	Feasibility analysis when deemed necessary; technical report required	Case by case	No	RES grants $2,500 to $1M; loan guarantees to $25M	25% to 50% depending on track
VAPG	Not mandatory but working-capital requests are scored on agreed financial feasibility	N/A	Yes, planning grants fund studies	Planning $50K, working capital $200K	~50%
RBDG	No	N/A	Yes, studies are an eligible use	Typically $10K to $500K	None
MPPEP	Yes	All applications	No	$10K to $2M by track	Varies
SBA 7(a) and 504	Lender or CDC discretion, common for new or special-purpose	Discretionary	No	To $5M (7a), $5.5M SBA portion (504 small manufacturer), $10M combined	Typically 10%+

A few notes for planning purposes. REAP grant awards were paused as of March 31, 2026 pending new regulations, while the guaranteed loan track continues. Most FY2027 windows are not yet posted, which makes the second half of 2026 the scoping season for studies that will support 2027 applications. And fertilizer manufacturing sits in NAICS 3253, inside the manufacturing sector, so it qualifies for SBA small-manufacturer treatment, though the $10 million combined ceiling makes SBA a partial-stack tool rather than a plant financing solution.

The through-line: the same five-dimension framework governs all of them. A study built properly for one is largely portable to the others, which is why the sequencing decision matters more than the program decision.

Part 11: What to do in the days you have left

If you are applying to FIELDS this cycle, the deadline is 11:59 p.m. Eastern on August 17, 2026, and the sequence is not negotiable.

1. Resolve your document set against Grants.gov, not rd.usda.gov. The $150 million ceiling, the pre-development cost restriction, and the removed packaging language only exist in the amended file.

2. Run the market share test and certify. Compute both global and domestic production-capacity shares against the tables in Part 6. For nearly every independent applicant this takes an afternoon and then it is done.

3. Confirm your match is documentable, not intended. Signed commitment letters, proof of funds, executed term sheets. Unsecured match is an eligibility failure, not a scoring deduction.

4. Get the off-take executed if it is anywhere close. It carries up to five bonus points directly and drives a large share of the 30-point market block.

5. Commission the study to lender standard, signed by an independent consultant with a documented qualifications statement. Build it to the seven elements and the five dimensions simultaneously so it serves FIELDS now and a B&I or conventional debt process later.

6. Underwrite the downside. Mid-cycle netbacks, forward-strip gas, 80 percent utilization, and an explicit sensitivity to a return to 2024 pricing.

If you cannot assemble the 50 percent match by mid-August, the disciplined move is to not apply this round. The Commodity Credit Corporation mechanism is repeatable and a rejected application is a worse position than a deferred one. Use the interval to build the evidence file, and use RBDG or a VAPG planning grant in a future cycle to fund the study itself where you are eligible.

The lesson underneath the 6 percent

The most important thing about the FPEP record is that the projects did not fail because the technology was fake or the sponsors were dishonest. They failed because nobody made them prove, before the money moved, that they had the other half of the capital, a buyer for the product, a supplier for the feedstock, a permit path, and a contractor's price rather than a spreadsheet's.

FIELDS turned every one of those into an application requirement. That is what the feasibility study is for. Not compliance theater, and not a business plan with a consultant's logo on it, but an independent opinion that survives a reviewer who has now read 121 projects' worth of reasons to be skeptical.

The projects that got built under FPEP were, almost without exception, the ones that could have proven all six things on day one. That is the whole finding.

Frequently asked questions

Does the FIELDS program require a feasibility study? Yes, for every application regardless of award size. The study must be prepared for the specific project, dated no more than three years before submission, signed by a qualified independent consultant, and the agency must concur that it is acceptable and adequate. This is a change from FPEP, which required a study only for grant requests of $5 million or more.

What is the FIELDS application deadline? 11:59 p.m. Eastern on August 17, 2026, per the notice, the Rural Development program page and the Grants.gov posting. USDA's July 1 press release stated August 15, which appears to be an error.

What is the maximum FIELDS award? $150 million, following the July 7, 2026 amendment on Grants.gov. The notice PDF hosted on rd.usda.gov still shows the original $100 million and is out of date. The minimum is $15 million.

Who counts as a qualified independent consultant for a USDA feasibility study? An independent third party with the knowledge, expertise and experience to perform the specific task, holding no financial interest in the project, the applicant or the outcome. USDA does not maintain an approved consultant list, and any claim of USDA pre-approval should be treated with suspicion. The study must include the author's resume or statement of qualifications.

What are the five dimensions of a USDA feasibility study? Economic, market, technical, financial and management feasibility, per 7 CFR 5001.3. The FIELDS study guide wraps these in an executive summary and a consultant recommendation, plus references, for seven elements total.

How much matching funding does FIELDS require? Fifty percent of total eligible project costs, from non-federal sources, secured and documented before the award agreement is executed. On a $50 million project, that means requesting up to $25 million and providing at least $25 million. Projects without secured match are ineligible.

Can a large fertilizer producer apply to FIELDS? Not if the applicant or its affiliates hold production market share at or above the fourth-largest producer in nitrogen, sulfur, phosphate or potash. In practice that threshold sits near 4.7 percent of U.S. ammonia capacity for nitrogen and is effectively non-binding in phosphate, potash and sulfur for a new entrant.

How many FIELDS awards will be made? Approximately ten, with awards anticipated in the December 2026 to January 2027 window.

Sources

1. U.S. Department of Agriculture, press release, December 18, 2024. FPEP investment of $517 million across 76 facilities in 34 states and Puerto Rico.

2. U.S. Department of Agriculture, "Secretary Rollins Announces $500 Million for Fertilizer Investment and Expansion Program," Release 0089.26, July 1, 2026.

3. DTN/Progressive Farmer, July 1, 2026, reporting on FIELDS launch and FPEP project status.

4. Red River Farm Network, May 20, 2026, reporting Secretary Rollins on unsatisfactory FPEP awards.

5. U.S. Department of Agriculture, Secretary Vilsack, March 2023, on FPEP application volume.

6. USDA Rural Development, Fertilizer Production Expansion Program Feasibility Study Guidelines.

7. DTN/Progressive Farmer, May 15, 2026, "Fertilizer Expansion Gains Momentum, But Finance, Competition Hurdles Remain."

8. AgWeb, July 2026, reporting Secretary Rollins on FPEP project composition.

9. Agri-Pulse, July 1, 2026, on FIELDS launch and FPEP figures.

10. USDA Rural Development, Notice of Funding Opportunity RD-RBS-26-01-FIELDS, published July 1, 2026, as amended July 7, 2026 on Grants.gov.

11. CF Industries Holdings, Form 10-K and Annual Report, FY2024 and FY2025, U.S. Securities and Exchange Commission.

12. Ammonia Energy Association, September 2025; LSB Industries Form 10-K, U.S. Securities and Exchange Commission.

13. U.S. Geological Survey, Mineral Commodity Summaries, nitrogen (ammonia), 2025 and 2026 editions.

14. The Mosaic Company, Form 10-K, U.S. Securities and Exchange Commission; company disclosures.

15. CRU / BC Insight, phosphate industry profile, May 2020.

16. The Fertilizer Institute; University of Illinois farmdoc, 2026.

17. 7 CFR 5001.3, Definitions, and Appendix A to Subpart D, published at 85 FR 42494, July 14, 2020.

18. U.S. Government Accountability Office, GAO/RCED-99-249, "Rural Development: Rural Business-Cooperative Service Business Loan Losses," August 25, 1999.

19. USDA Office of the Chief Counsel published application throughput data, FY2021 to FY2023.

20. DTN retail fertilizer price index, March through July 2026.

21. World Trade Organization Data Blog, July 10, 2026, on fertilizer trade disruption and pricing.

22. CNBC and global-agriculture.com, March 2026, on Qatari urea production halt.

23. U.S. Geological Survey, Final 2025 List of Critical Minerals, Federal Register, November 7, 2025.

24. Texas A&M Agricultural and Food Policy Center, January 2026, via DTN.

25. U.S. Department of Justice, Antitrust Division, press release, September 26, 2025.

26. USDA Economic Research Service, Farm Sector Income Forecast, February 5, 2026.