Worked Example: SBA 7(a) Gas Station Feasibility Study in Tennessee

This is a worked example, not a client engagement. It takes a ground-up fueling station and convenience store in unincorporated Wilson County, Tennessee, assumes a single SBA 7(a) loan for the whole project, and walks through what a feasibility study would have to show a lender before that loan could be sized. Every market, program and tax figure is sourced and cited in the sentence that uses it; every project figure that is not sourced is labelled an assumption. The example is written for a borrower planning a first store, for the SBA lender who would underwrite it, and for anyone deciding whether a feasibility consultant would earn a place in the file.

At a glance. An assumed 4,000 square foot store with a foodservice counter and ten fueling positions at a Wilson County interchange, financed with one SBA 7(a) loan at prime plus an assumed 2.75 point spread. At the assumed stabilized volume the base case covers debt service 1.04x, which is below both the 1.15x program floor and that lender minimum; the question the study would have to answer is whether the site, the margin or the structure can close that gap.

• Total project cost: $4,727,215
• SBA 7(a) loan: $3,545,411
• Borrower equity: $1,181,804
• Stabilized EBITDA available for debt service: $426,162
• Debt service coverage at stabilization: 1.04x
• Break-even volume at 1.00x (percent of base case): 98.6%
• Volume needed for the lender's 1.25x minimum (percent of base case): 107.0%

What a Feasibility Consultant Tests First: Can the Site Carry the Debt?

A lender looking at a ground-up fueling station is asking whether this corner, with these tanks and this store, will sell enough gallons and enough sandwiches to pay the note every month for 25 years, starting in a first year when nobody yet knows the store exists.

The SBA's credit standard puts the question in a few words: a loan must be "so sound as to reasonably assure repayment" ^[1]. The same regulation lets a lender weigh the applicant's earnings or cash flow, its credit history and its equity or collateral, and it sets no numeric coverage minimum of its own ^[2]. This example did not read the SOP itself. Coverage is therefore tested against two minimums: the 1.15x floor on business coverage that the site's own SBA program page states for standard 7(a) loans ^[3], and the 1.25x debt service coverage at stabilization that this example assumes a lender would insist on; the higher of the two binds, and here that is the assumed 1.25x. A lender can also ask for a feasibility study as a loan condition ^[4], and for a start-up with no operating history a feasibility study is the main evidence of future cash flow a file can hold.

So the consultant's first test is not the market; it is the arithmetic the market has to satisfy. In this example the debt service and coverage table below prints total annual debt service of $409,565 against stabilized EBITDA available for debt service of $426,162, a coverage of 1.04x. That sits under the 1.15x floor and under the assumed 1.25x, and the table prints the higher, 1.25x, as the binding minimum.

Decision. Coverage, not loan to cost, would set the loan, because the binding minimum cuts in before the cost ratio does: the assumed 75 percent loan to cost gives $3,545,411, while base earnings carry only $2,951,266 at the assumed 1.25x minimum that binds (the debt service and coverage table).

Three tests follow from that: volume, where the deciding assumption is 3,500 gallons of gasoline and 450 gallons of diesel a day; margin, where the model assumes 36.0 cents a gallon before card fees; and structure, meaning the equity, the rate and any reserve for the ramp years.

The Site and Its Market in Tennessee

The example places the project on an assumed parcel of about 1.75 acres, carried at $700,000, at a highway interchange in unincorporated Wilson County, on the eastern edge of the Nashville metropolitan area, in a county crossed by Interstate 40 and Interstate 840. No city, exit or road name is given for the site itself, and the choice of an unincorporated location is deliberate: Tennessee applies the county or city tax rate per $100 of assessed value to commercial real property assessed at 40 percent of appraised value ^[5], and the 2025 Tax Aggregate Report lists Wilson County's total county rate at $1.9089 per $100 of assessed value, an effective 0.7636 percent of appraised value for commercial real property ^[6]. A site inside a city would carry a municipal rate on top of the county rate.

Why Wilson County rather than somewhere else in the state? Population. Wilson County held 175,033 people at the July 1, 2025 mark in Census Vintage 2025, compared with 147,745 in the April 1, 2020 base, a gain of about 18.5 percent, with 4,693 added in the single year to July 2025 ^[7]. Tennessee as a whole grew about 5.8 percent over the same period, to 7,315,076 ^[8], and the nation about 3.1 percent, to 341,784,857 ^[9].

The Census Bureau's reading of the Vintage 2025 numbers fits the site. Growth in metro areas since 2020 has largely been in the exurban areas on their outer edges ^[10], and among county types it was the metro outlying counties in the South that grew the fastest ^[11].

What a consultant cannot say from the public record is what a lender will want most: a traffic count at the interchange and the household income of the trade area. The Tennessee Department of Transportation publishes traffic counts, but its pages could not be retrieved for this example, so no annual average daily traffic figure is quoted and none is modelled. The study would fill that gap with a field or purchased count.

Decision. The study would treat the site's volume as unproven until a traffic count and a trade-area income figure are in hand; the population figures support the choice of county, not the gallons per day.

Demand: Fuel, Traffic and Population in Wilson County

Fuel. Table MF-21 of the Federal Highway Administration puts Tennessee's 2024 highway motor-fuel use at 3,203,686 thousand gallons of gasoline plus 1,136,182 thousand gallons of diesel and other special fuels, about 4.3 billion gallons in all ^[12]. Against that state total the project's assumed 1,441,750 annual gallons is a rounding error: a single store competes for the share of one corridor that will turn in at one exit.

Roads. FHWA's Table VM-2 reports 78,904 million vehicle-miles traveled in Tennessee in 2024, of which 9,838 million were on rural interstates and 16,138 million on urban interstates, 25,976 million on interstates in all ^[13]. The assumed site sits where interstate and local traffic meet, which matters for the mix the model assumes: 3,500 gallons of gasoline a day against 450 of diesel, with diesel at every dispenser but no truck island, so the diesel is light-truck and automotive diesel rather than over-the-road trucking, and the store does not compete with travel centers for the truck stop.

Benchmark. For 2025 the NACS State of the Industry figures give the typical convenience store 322 fuel sales daily at 8.4 gallons apiece, roughly 2,705 gallons a day ^[14]. The model's assumed 3,950 gallons a day (3,500 gasoline and 450 diesel) is about 1.46 times that average, and the inputs table says so in its basis column. Nothing in the sourced record proves a Wilson County interchange store does that volume; the figure is an assumption about the site, chosen because a ten-position canopy at an interstate interchange that sold only the average store's gallons would not justify its fuel system. The demand table below gathers the sourced indicators in one place.

Decision. The study would carry the state fuel and traffic figures as context only and would rest the gallons-per-day assumption on a site count it does not yet have; until then the volume is the weakest number in the model.

Competition: Price, Fueling Positions and a Crowded Store Count

A new fueling station in Tennessee enters a market that is already dense. The 2026 NACS store count, built on NIQ TDLinx data, found 122,620 convenience stores selling motor fuel in the United States as of December 2025, 80.7 percent of the 151,975 convenience stores in the country ^[15]. Those stores averaged 9.8 fueling positions each, according to the NACS State of the Industry Report on 2024 data ^[16]; the example assumes 10 positions, five dispensers under one canopy, a hair above the average and sized to the volume the model assumes rather than to the lot.

Competition at the pump is price competition. In the NACS May 2025 Consumer Fuels Survey, 72 percent of drivers said price is the most important factor in choosing where to buy fuel, against 16 percent who cited location and 12 percent brand ^[17]. For a new entrant with no loyalty base that number sets the margin strategy: a store that must be visibly cheaper than the canopy across the interchange cannot also take the industry's full margin. That is why the model assumes a fuel gross margin of 36.0 cents a gallon before card fees, below the OPIS figure NACS cites of 39.7 cents in 2025 (12.7 percent of the $3.12 average price for the year) and below the five-year average of 38.3 cents ^[18].

A margin of that size is thin once the cost of selling fuel is taken out. NACS estimates that retailer expenses to sell fuel run about 20 to 25 cents a gallon, about 8 cents of it credit card fees ^[19]. Card fees scale with the pump price, not the gallon, and the pump price is high: EIA's weekly survey put the Midwest (PADD 2) average for regular gasoline at $4.102 a gallon on November 2, 2026, $1.274 above a year earlier ^[20], and on-highway diesel at $6.402, $2.629 above a year earlier ^[21]. Tennessee sits in PADD 2 ^[22], and the EIA prices include all taxes ^[23], so they are the prices the model carries at the pump. Inside those prices sits the state fuel tax, which FHWA's Table MF-205 shows at 26 cents a gallon on gasoline ^[24] and 27 cents on diesel ^[25] as of December 31, 2024 and each of the five year-ends before it; the retailer collects it and keeps none of it.

The inside of the store competes on a different axis. NACS reports that fuel was 65.0 percent of convenience store sales dollars in 2025 but only 38.8 percent of gross profit dollars ^[26], and that foodservice alone was 28.5 percent of in-store sales and 38.9 percent of in-store gross margin dollars ^[27]. The store that wins the fuel customer on price has to earn its profit inside, and inside it has to earn it at the foodservice counter. The example's assumed 4,000 square foot store with a made-to-order counter sits inside the 4,000 to 5,000 square foot range that NACS classes as the hyper convenience format, stores that offer products and services by department ^[28].

Competitors that matter here are not only other fueling stations. The real estate itself trades in a market that is softening: Matthews Real Estate Investment Services reported in March 2026 that net lease convenience store listings had risen roughly 27 percent since mid-2025 while cap rates had begun drifting upward and pricing dispersion had widened ^[29]. Matthews gave no cap rate figure, so this example quotes none, but a lender would ask whether the collateral behind a 25-year term is worth on day one what it cost.

Decision. The study would model the project as a price follower, not a price leader: the assumed 36.0 cent margin is held below the OPIS figure NACS cites ^[18] on purpose, and the sensitivity table tests what a further 5 cent cut does to coverage.

SBA Lending Record for This Asset Class in Tennessee

Loan Analytics keeps a loan-outcomes dataset built from the SBA's own loan-level 7(a) and 504 files ^[30]. For this example's state and asset class it covers SBA loans approved in fiscal 2017 to 2026 to businesses coded to gas station and truck stop, with the project in Tennessee. The figures are the SBA program's record in the state, not Loan Analytics' work and not this hypothetical project. The dataset notes a limit on the class: the code is the borrower's industry. Truck stop or travel center cannot be split out. EV charging has no entry in the 2022 NAICS index.

Cost Basis: What the Project Would Cost and Who Would Fund It

The inputs table below is the whole model in one place, with each line's basis beside it.

Rounded to the dollar; lines may not foot by $1.

The cost side is built from the ground up. The assumed land cost of $700,000 for about 1.75 acres reflects an interchange parcel in unincorporated Wilson County; no public land price was sourced. Site work, paving and utilities are assumed at $550,000, a large line for a fueling station because the canopy, the tank field, the turning radii and the stormwater all sit on the pad rather than in the building. The assumed building hard cost is $300 a square foot on the assumed 4,000 square feet, $1,200,000 in all, for a store with a commercial kitchen, walk-in coolers and a seating area.

The fuel system is the line that separates this asset class from every other kind of retail. The model assumes $850,000 for the canopy, five dispensers, three double-wall underground tanks, piping, sumps and monitoring. Under 40 CFR 280.20, tanks and piping installed or replaced after April 11, 2016 must be secondarily contained and use interstitial monitoring ^[31]; there is no cheaper single-wall option for a new build. Tennessee administers and enforces its own petroleum underground storage tank program in lieu of the federal one, through the Tennessee Department of Environment and Conservation ^[50], a program EPA approved effective January 15, 1999 and revised effective November 8, 2021 ^[51]. The state's own rules could not be retrieved for this example, so the model carries the federal standard and an assumed $12,000 a year for licenses, permits, tank fees and compliance testing, with the Tennessee fee schedule noted as unsourced.

Store and foodservice equipment, fixtures and point of sale are assumed at $450,000; architecture, engineering, permits and impact fees at $180,000; interest during construction at $130,000, about half the loan drawn for nine months at the note rate; and a combined $95,000 line for third-party reports, closing and financing costs, not broken out further. An assumed construction contingency of 5 percent of site, building and fuel system adds $130,000.

Two lines would not appear in a 504 project at all. Opening inventory, merchandise and the first fuel loads, is assumed at $220,000, and pre-opening costs and working capital at $125,000. 7(a) proceeds may fund inventory, supplies, raw materials and working capital ^[32] on top of land, site improvements, new construction and fixed assets ^[52]; the program section explains why that matters for the choice of program.

The cost table adds the SBA guaranty fee to the project. The model assumes the maximum guaranty, which for a 7(a) loan of more than $150,000 is 75 percent ^[35], and takes the guaranty fee at its ceiling: 3.5 percent at most of the guaranteed part when the loan tops $700,000 ^[36], plus a further 0.25 percent that reaches only the guaranteed dollars past $1,000,000 ^[38]. The two texts differ on the addition, the regulation applying it to the guaranteed portion at large ^[36] and the statute reaching only dollars past the $1,000,000 mark ^[38]; this example prices the fee at the statutory ceiling. On a guaranteed portion of $2,659,058 that makes the fee $97,215, and the borrower may pay it from loan proceeds ^[37], which is how the table treats it. The fiscal-year fee notice may set a lower fee; this example did not retrieve it and does not cite it.

With the fee added, total project cost is $4,727,215, funded by an SBA 7(a) loan of $3,545,411 at the assumed 75 percent loan to cost and borrower equity of $1,181,804, a 25 percent share. The loan sits well inside the $5,000,000 maximum for any one 7(a) loan ^[53]. A larger store on a larger parcel was the starting concept for this example and was set aside: the cost moved toward the loan cap at a sensible equity share, and coverage was already thin at the smaller size, so the example is the smaller store.

Decision. The study would carry the fuel system and the UST compliance lines at the federal secondary containment standard ^[31] and flag the Tennessee fee schedule as unsourced, rather than assume the state adds nothing.

Revenue Build: Gallons, Inside Sales and the Ramp

The revenue table starts at the pump. The assumed 3,500 gallons of gasoline and 450 gallons of diesel a day, 1,441,750 gallons a year, are sold at the EIA PADD 2 prices of $4.102 for regular ^[20] and $6.402 for diesel ^[21], giving fuel sales of $6,291,834. Fuel sales are the largest number on the page and the least important: at the assumed margin of 36.0 cents a gallon they produce $519,030 of gross profit, 35.8% of the total, against the 38.8 percent of gross profit dollars NACS reports fuel earned across the industry in 2025 ^[26]. The project leans harder on the inside than the industry does, by design.

Inside sales are assumed at $2,550,000: $1,850,000 of merchandise excluding foodservice at an assumed 30.0% gross margin, and $700,000 of foodservice at an assumed 48.0% margin. The merchandise total puts inside sales at about 1.14 times the NACS 2025 in-store sales per store, that is $341.2 billion of in-store sales ^[39] spread over the 151,975 stores NACS and NIQ count ^[15], a ratio the inputs table notes as the basis of the assumption. The foodservice share, 27.5 percent of inside sales, sits just below the 28.5 percent NACS reports for the industry ^[27], and the assumed 48.0% foodservice margin is 1.6 times the assumed merchandise margin, the ratio implied by foodservice's 28.5 percent of inside sales and 38.9 percent of inside gross margin dollars in the same NACS figures ^[27]. Other income from lottery commissions, the ATM and air and vacuum is assumed at $40,000; no lottery commission rate was sourced.

Inside gross margin comes to 34.9% and total gross profit to $1,450,030 on total revenue of $8,881,834.

No new store opens at its stabilized volume. The ramp table assumes 80% of base volume in year 1 and 92% in year 2, scaling gallons, inside sales and other income together, before the base case arrives in year 3. In year 1 that is 1,153,400 gallons and $2,040,000 of inside sales, and EBITDA available for debt service of $182,133 against annual debt service of $409,565, a coverage of 0.44x. Year 2 improves to $328,550 of EBITDA and 0.80x. In both ramp years debt service exceeds EBITDA, and the gap is far larger than the $125,000 of pre-opening costs and working capital the cost basis assumes. That shortfall has to be funded by more working capital in the loan, a funded interest reserve, or the borrower's own liquidity, and the lender would want to see which before the loan is sized.

Decision. The study would present year 3 as the stabilized year and would not average the ramp into it; the lender's question is whether the borrower gets through years 1 and 2, which the ramp table answers in the negative at the assumed volumes unless a reserve is added.

Debt Service Coverage and Sensitivity

The stabilized operating statement takes $1,450,030 of gross profit and spends $1,023,868 of it. The largest line is store payroll, taxes and benefits at $443,256, built from an assumed 64 labor hours a day for a 24-hour store with two on shift most hours and foodservice coverage, an assumed average wage of $16.50 an hour, above the NACS 2025 store-level average of $15.04 ^[42], and an assumed 15 percent for payroll taxes, workers compensation and benefits. The second is card fees at $229,888, an assumed 2.6 percent of fuel and inside sales, the ratio of the record $21.3 billion in card fees NACS reports convenience retailers paid in 2025 ^[40] to the $817.5 billion of total industry sales NACS reports for the year ^[41].

Card fees are the expense the operator controls least: they follow the pump price, and the PADD 2 price was $1.274 a gallon higher for gasoline ^[20] and $2.629 higher for diesel ^[21] than a year earlier. NACS research staff reported in August 2026 that direct store operating expenses have risen sharply over the past five years and continue to outpace pretax profit ^[54].

The tax lines are the model's most fully sourced expenses. Real property tax of $24,434 applies Wilson County's 2025 rate of $1.9089 per $100 of assessed value ^[6] to 40 percent of an assumed $3,200,000 appraised value ^[5]; tangible personal property tax of $2,291 applies the same rate to 30 percent of an assumed $400,000 of equipment ^[45]. Insurance, including the tank pollution liability a petroleum marketing facility must carry, is assumed at $40,000; the financial responsibility floor under 40 CFR 280.93 is $1 million per occurrence ^[43] and $1 million in annual aggregate for an owner of 1 to 100 tanks ^[44]. Utilities at $72,000, repairs and maintenance at $36,000, accounting, legal and technology at $24,000, advertising and loyalty at $15,000, supplies at $30,000, an owner-manager salary of $70,000 and a replacement reserve of $25,000 are all assumptions, and the inputs table marks each one.

EBITDA available for debt service comes to $426,162, 4.8% of total revenue.

The debt table prices the loan at prime plus the assumed spread. Every daily reading of prime in the week of November 2, 2026 was 7.00 percent ^[46]. A floating 7(a) rate can be built on prime or on the Optional Peg Rate ^[55], and for a loan above $350,000 it is capped at 3.0 percentage points over the base rate ^[47]. The assumed spread of 2.75 points sits inside that cap and gives a note rate of 9.75%.

The loan is divided by purpose. Real estate, and equipment with a useful life beyond ten years, escape the ten-year ceiling on a 7(a) term ^[49]; the part spent on buying or improving real property can run 25 years, with extra months allowed for finishing construction, though the 25-year maximum still caps the whole ^[48]. The model puts the real estate share, $2,936,642, on 25 years and the equipment, inventory and working capital share, $608,769, on ten years, for annual debt service of $314,034 and $95,531 respectively, $409,565 in all, a loan constant of 11.55%.

Coverage is 1.04x, below the 1.15x floor and below the assumed 1.25x lender minimum. The higher of the two binds, so the table's binding minimum is the assumed 1.25x, and the loan that base earnings can carry at that minimum is $2,951,266. On that smaller loan the guaranty fee shrinks, and the table prints the SBA guaranty fee recomputed on that loan, $80,504, the total project cost at that loan, $4,710,504, and the equity at that loan, $1,759,238, against $1,181,804 at the assumed 75 percent loan.

The sensitivity table moves one input at a time. A ten percent fall in volume takes coverage to 0.74x and a 20 percent fall to 0.44x; two points on prime take it to 0.91x; a ten percent rise in operating expenses to 0.79x; and 5 cents off the fuel margin to 0.86x. No single stress leaves the project above 1.00x, and none of them is extreme: a new store that opens into a price response from the canopy across the interchange could see the volume case and the margin case at once.

The break-even table turns the same arithmetic around. At 1.00x coverage the project needs 98.6% of its base volume, 3,896 gallons a day and $2,515,314 of inside sales; at the 1.15x program floor it needs 103.7%, 4,095 gallons a day and $2,643,708 of inside sales; at the assumed 1.25x binding minimum it needs 107.0%, 4,228 gallons a day and $2,729,303 of inside sales. On fuel margin alone, holding volume at base, the project covers debt service at 34.8 cents a gallon, reaches the floor at 39.1 cents and reaches the binding minimum at 42.0 cents; the model's assumed 36.0 cents sits between the first two. The OPIS figure NACS cites for 2025, 39.7 cents ^[18], would clear the floor and still fall short of the binding minimum, and the five-year average NACS cites from OPIS, 38.3 cents ^[18], would lift coverage above 1.00x and still leave it under the floor.

Decision. The study would report the base case as short of both the floor and the assumed 1.25x binding minimum and would not present an upside case as the base; the figures a lender would act on are the loan the base case supports, $2,951,266, and the break-even margin and volume at the binding minimum, 42.0 cents and 4,228 gallons a day.

What the SBA 7(a) Rule Requires and How the Study Would Meet It

Why 7(a) and not 504? Because of what the project needs to buy. Proceeds of any SBA business loan may be used to acquire land, improve a site, construct new buildings and acquire and install fixed assets ^[52], and 7(a) proceeds may also fund inventory, supplies, raw materials and working capital ^[32]. A 504 loan cannot cover ancillary business expenses like working capital, franchise fees or advertising ^[56]. This project carries an assumed $220,000 of opening inventory and $125,000 of pre-opening costs and working capital, and the ramp table shows it needs every dollar of them; under 504 those lines would have to come from equity or a separate loan.

Equity is the second reason. 13 CFR 120 sets no equity floor for a 7(a) loan, and the model's 75 percent loan to cost is an assumption of lender practice. On the 504 side, two conditions each raise the borrower's contribution under 13 CFR 120.910(a): a limited or single purpose building, which this example takes a fueling station to be ^[33], and a business in operation for no more than two years ^[34]. Each carries 15 percent, and with both present, as for a start-up fueling station, the floor is 20 percent ^{[33] [34]}. So 504 would ask for 20 percent before the first dollar of working capital, which it could not fund anyway.

504 could still lower the blended rate on the real estate share. No debenture rate was sourced for this example, so the 504 alternative is named and not modelled.

The loan cap, the guaranty, the fee, the rate cap and the terms are covered where the cost and debt tables use them; all of them sit inside 13 CFR 120 as cited there, and the loan is well inside the $5,000,000 maximum ^[53].

Occupancy. Where SBA finances new construction, the borrower has to use 60 percent or more of the rentable space itself, and up to 20 percent may be leased out permanently ^[57]. The operator here occupies the whole building; there is no tenant space in the assumed 4,000 square feet, so the rule is met with room to spare. Passive ownership is not an issue either: the business sells fuel and food from the asset it owns, which is the opposite of the landlord case 13 CFR 120.110(c) excludes ^[58].

Credit. The regulation requires a loan "so sound as to reasonably assure repayment" ^[1]; the lender may weigh the applicant's earnings or cash flow, credit history and equity or collateral, without setting a numeric coverage minimum ^[2]. SBA may require a feasibility study as a condition ^[4]. The study would meet that requirement with a cost basis every line of which is sourced or marked, a revenue build tied to industry benchmarks and a site count, and a coverage test against the 1.15x floor and the assumed 1.25x lender minimum, reported plainly when the base case misses them. The program statements here rest on 13 CFR 120, apart from the coverage floor, which is cited to the site's own SBA page ^[3].

Decision. The study would recommend 7(a) over 504 on the facts of the use of proceeds ^[56] and the equity floor ^[33], and would present a 504 split only if a debenture rate were sourced.

Risks a Lender Would Price

Volume on a new site. The deciding assumption is 3,500 gallons of gasoline and 450 of diesel a day, about 1.46 times the NACS average of 322 transactions at 8.4 gallons ^[14]. No traffic count supports it. The sensitivity table shows a 20 percent miss takes coverage to 0.44x.

Price competition and margin compression. 72 percent of drivers choose on price, NACS found ^[17]. The model already assumes a 36.0 cent margin below the 39.7 cents OPIS data cited by NACS show for 2025 ^[18]; the break-even table puts 1.00x at 34.8 cents. A sustained price war is a solvency risk, not a profit risk.

Card fees and expense growth. NACS reports card fees at a record $21.3 billion in 2025 ^[40], store wages averaging $15.04 an hour ^[42], and direct store operating expenses outpacing pretax profit over five years ^[54]. The sensitivity table's ten percent expense case prints 0.79x.

Interest rate. The note is variable, priced off prime ^[55] under a 3.0 point cap ^[47]. Two points on prime take coverage to 0.91x in the sensitivity table, and the ramp years would carry whatever prime does then.

Fuel price swings. High pump prices ^[20] lift card fees and the cost of the first fuel loads in the assumed $220,000 of opening inventory; distribution and marketing were 12.5 percent of a $4.058 national gallon in August 2026 ^[59], and the retailer's margin is a slice of that slice.

Environmental and tank compliance. Secondary containment and interstitial monitoring are mandatory for new tanks ^[31]; financial responsibility of $1 million per occurrence is mandatory for a petroleum marketing facility ^[43]; and Tennessee's own program, run by TDEC in lieu of the federal one ^[50], may add fees and rules this example could not retrieve. A release is both a cleanup liability and a loss of the collateral's value.

Collateral value. Matthews reported in March 2026 that cap rates on net lease convenience stores had begun drifting upward and pricing dispersion had widened as listings rose roughly 27 percent since mid-2025 ^[29]. A lender on a 25-year real estate term would discount the completed value, not take cost as value.

Ramp-up losses. The ramp table prints 0.44x and 0.80x in years 1 and 2 at the assumed 80% and 92% volume indices; the assumed $125,000 of working capital does not cover the gap.

Electric vehicle adoption. No figure was sourced for this example, so none is quoted; the study would have to source one for the corridor and for the horizon of a 25-year term.

Decision. The study would rank volume first, margin second and structure third, because the sensitivity table shows volume and margin can each take coverage below 1.00x on their own, while structure is the only one of the three the borrower controls.

What Would Decide the Outcome

The base case misses the floor and the binding minimum. A study that stopped there would be useless to the borrower, and one that papered over it would be useless to the lender.

The restructure. The restructure table takes the $2,951,266 that base EBITDA carries at the assumed 1.25x binding minimum and rounds it down to a 7(a) loan of $2,950,000; every other cost line, interest during construction included, stays as assumed, and only the guaranty fee moves, recomputed under the statutory rule at $80,469. Total project cost becomes $4,710,469 and borrower equity $1,760,469, 37.4% of that cost and $578,665 above the base case. Debt service falls to $340,783 and coverage reaches 1.25x, clearing the 1.15x program floor and meeting the 1.25x assumed for the lender: the restructure passes both tests, though by only $183 of EBITDA. It does not carry the ramp years: at that debt service years 1 and 2 still fall short by $170,883, so the borrower would also hold a $45,883 reserve beyond the assumed $125,000 pre-opening and working capital line. With that equity and that reserve, the project is financeable at the assumed volume and margin with no change to the site. If not, the question moves to the site.

The volume the site would need. The break-even table prints 4,228 gallons a day for the binding minimum, against the 3,950 already assumed and the NACS average of about 2,705 ^[14]. A traffic count and a turn-in rate are the only evidence that would settle whether 4,228 is plausible; without them the study would say so.

The margin the site would need. The break-even table prints 42.0 cents a gallon for the binding minimum at base volume, above both the 39.7 cent 2025 margin and the 38.3 cent five-year average that OPIS data cited by NACS show ^[18]. A new entrant pricing to win share, in a market where 72 percent of drivers choose on price in the NACS survey ^[17], is unlikely to run above the industry average in its first years. The margin route is the least credible of the three.

The structure. A funded reserve for the ramp years, sized from the ramp table's year 1 and year 2 EBITDA of $182,133 and $328,550 against debt service of $409,565, would address survival but not stabilized coverage. A 504 split on the real estate could lower the blended rate, but no debenture rate was sourced and the equity floor would be 20 percent for a start-up in a single purpose building ^{[33] [34]}. More equity is the structure that works on the model's own numbers.

Decision. The study would size the 7(a) loan at the restructure's $2,950,000, not $3,545,411, and would hand the borrower three ways to close the gap in order of credibility: equity, then a proven site volume above 4,228 gallons a day, then margin.

Questions to Ask a Feasibility Study Company

Does the base case have to meet the lender's minimum for the study to be useful? No. A study is a test, not an advertisement. In this example the base case prints 1.04x against a 1.15x floor and an assumed, binding 1.25x minimum, and the useful output is the restructure it points to, a $2,950,000 loan carrying $1,760,469 of equity. A lender can work with that; a lender cannot work with a base case that was tuned until it passed.

Why 7(a) rather than 504 for a gas station? Because the project needs inventory and working capital, which 7(a) proceeds may fund ^[32] and 504 proceeds may not ^[56], and because a start-up in a single purpose building faces a 20 percent contribution floor under 504 ^{[33] [34]}. 504 can still lower the rate on the real estate share, and a study that compared the two would source the debenture rate first.

Why does the model use no Tennessee traffic count and no Tennessee tank fee? Because neither could be retrieved from the state publisher for this example. The fuel taxes quoted, 26 cents a gallon on gasoline ^[24] and 27 cents on diesel ^[25] as of December 31, 2024, come from FHWA's table of state rates, and the tank rules come from 40 CFR 280 ^[31]; Tennessee's own UST program ^[50] may add requirements the example does not state.

What makes a gas station feasibility consultant different from a generalist? Fuel. A generalist models a store; a fuel station consultant models a tank field, a card fee that moves with crude, and a margin that 72 percent of customers shop on price for, as the NACS survey shows ^[17]. Ask whether the consultant has run a break-even in cents per gallon, as the break-even table does here, and whether the sensitivity cases include a margin cut, not just a revenue cut.

What would a lender want added before relying on this? A traffic count at the interchange, a trade-area household income figure, a competitor canopy survey with posted prices, a Phase I environmental report, and bids for the fuel system and the building in place of the assumed $850,000 and $300 a square foot. None of those were available from public sources for this example.