Worked Example: USDA B&I Gas Station Acquisition and Expansion Feasibility Study in Pennsylvania

This is a worked example, not a client engagement. It tests a hypothetical purchase of an operating fuel stop sited in a rural-eligible area of Pennsylvania, together with a small diesel and truck-parking expansion, funded through a guaranteed loan under the USDA B&I program. It is written for a lender asking whether an acquisition-plus-expansion story survives its arithmetic, and for a buyer who wants to see which assumptions carry the result. The base case passes both tests it is held to, but thinly, and the pages below say how thinly and what would break it.

At a glance. A buyer purchases an existing fuel stop at the low end of a broker's published price range and adds one high-flow diesel island and the minimum truck parking on adjoining land, using an 85 percent USDA guarantee. The seller's own last fiscal year covers the proposed debt service just above the lender minimum, and the study's base case covers it comfortably more; the sponsor's own pro forma shows more again.

• Total project cost: $2,512,254
• USDA B&I guaranteed loan requested (85% guaranteed): $1,810,000
• Sponsor equity, share of total project cost: 28.0%
• Sponsor's pro forma coverage (market figures): 2.07x
• Study base case stabilized coverage: 1.59x
• Historical earnings coverage on the requested loan: 1.26x
• Volume needed for the 1.25x lender minimum (percent of base case): 94.3%

What a Feasibility Consultant Tests First on a Fuel-Stop Acquisition

A lender reading this request would ask one question before any other: do the earnings the buyer inherits carry the debt the purchase creates, before any credit is given to the expansion? The freight narrative, the truck-parking shortage and the fuel-tax regime matter only through that question. A consultant would therefore split the project into the part that exists and the part that must be built, and test the existing part alone against the full proposed debt service.

The existing part is an operating stop with eight fueling positions. In the example it sells an assumed 2,500 gallons of gasoline and 1,250 gallons of diesel a day, with assumed merchandise sales of $1,250,000, foodservice sales of $420,000 and other income of $35,000. Those assumptions stand for the seller's record, since no statement exists here; a real study would replace each with tax returns and monthly point-of-sale data. Read as a record, they give earnings after a paid manager and a replacement reserve of $232,757, as Table 12 prints.

The new part is one high-flow diesel island (two added positions), a diesel tank of an assumed 30,000 gallons, and the 20 truck spaces that Pennsylvania's truck stop test names as its minimum. The cost table, Table 7, prints a total project cost of $2,512,254, a guaranteed loan of $1,810,000 and sponsor equity of $702,254, which is 28.0 percent of cost. Annual debt service is $185,231. The seller's last year covers it 1.26x, the study's base case at stabilization 1.59x, and the sponsor's pro forma on market figures 2.07x, against an assumed lender minimum of 1.25x.

Two features keep this a close call: the seller's history, not the expansion, sets the request, and the price sits at the low end of a broker's range, which is what lets the equity fit.

Decision. The consultant would size the loan on the seller's history at the assumed lender minimum, show the expansion as upside the history does not need, and stress-test the thin margin over 1.25x rather than treat it as comfort.

The Market in Pennsylvania and the Mid-Atlantic Region

Pennsylvania is not a growth market for residents. Census Vintage 2025 estimates put the state at 13,059,432 people on July 1, 2025, against an April 1, 2020 estimates base of 13,002,753, about 0.4 percent growth in five years ^[1]. A fuel stop here does not live on new households. It lives on through traffic, and on the share of that traffic that is heavy.

On that point the federal highway statistics are specific. FHWA Table VM-4 puts combination trucks at 26.48 percent of vehicle-miles on Pennsylvania rural interstates in 2024 ^[2], and Table VM-2 reports 10,487 million vehicle-miles on those interstates, of 34,901 million on all rural roads and 95,194 million in total ^[3]. Table MF-21 shows 2024 highway use in Pennsylvania of 4,231,228 thousand gallons of gasoline and 1,577,089 thousand gallons of special fuels, mainly diesel ^[4]. Diesel is the smaller pool, and the stop's share of it is what the expansion turns on.

Prices frame the revenue. EIA put the Central Atlantic (PADD 1B) average retail price of regular gasoline at $4.146 per gallon on November 9, 2026, against $3.995 for the United States ^[5], and on-highway diesel at $6.412 against $6.127 ^[6]. Pennsylvania sits in PADD 1B ^[7] and the series includes all taxes ^[8], so the revenue lines in Table 8 are tax-inclusive.

Rural has two meanings here. The Center for Rural Pennsylvania uses its own density test, counting 48 rural and 19 urban counties, with nearly 3.4 million residents (26 percent) in the rural ones in 2020 ^[13]; the example does not assume that settles USDA eligibility. For a B&I guarantee, rural leaves out any municipality larger than 50,000 residents along with the developed land joined to it, and counts the rest ^[14]. The example is sited in a rural-eligible area as that rule defines it; a real study would confirm the parcel on the Agency's own eligibility map, which this example did not fetch.

On the transaction side, Matthews Real Estate Investment Services described net lease convenience store listings in March 2026 as up roughly 27 percent since mid-2025, with cap rates drifting upward and pricing dispersion widening ^[15]. That is a national statement about net-leased stores, but dispersed pricing is why a price needs a stated basis.

Decision. The market case rests on corridor freight, not population.

Supply and Public Comps: Truck Parking, Fuel-Stop Competition and Store Prices

No business is a party here; the public record offers counts, averages and one listed retailer's annual report, not transactions at this site.

FHWA's Jason's Law survey counted in Pennsylvania 66 public truck parking facilities with 1,569 spaces and 223 private truck stops with 9,363 spaces, 10,932 in all ^[16]. Private operators reported an average of 113 spaces per facility, and a majority reported a need for more ^[17]. FHWA reported that Pennsylvania has high levels of truck parking yet shortages at private truck stops ^[18], and surveyed drivers and trucking management cited the Mid-Atlantic as a region with shortages ^[19].

The shortage supports the expansion. The count also limits it: the 223 private stops hold most of the spaces, and the model adds only the 20 the state's test names, far below the 113 the average private operator reported. The diesel per added position is therefore assumed at 250 gallons a day, below the 276 the NACS averages imply: a new island must win trucks on price from stops with showers and restaurants.

A 2012 NATSO article quoted a directory publisher on chain truck stops having 75 to 125 parking spaces, among about 275,000 nationally ^[20], and added that, of 6,000 listed locations, about 3,500 are a convenience store with a pump and no showers or parking ^[21]. The second figure describes the format this project is closest to.

Fuel retailing is price-led. Price came first for 72 percent of drivers when NACS asked in May 2025 where they choose to buy fuel; 16 percent put location first and 12 percent brand ^[22]. NACS data on diesel give no margin anchor, and EIA's refiner diesel series by sales type reports no data after 2021, so the diesel margin is an assumed 30.0 cents, a judgement.

The retailer is Casey's General Stores, a listed Midwest chain, and the figures come from its Form 10-K. Casey's measures average store operating income as retail sales less cost of goods sold, store operating expenses and store depreciation and amortization, before interest, income taxes and company expenses not attributable to a store ^[26]. Casey's reports acquiring 40 stores in fiscal 2026 for payments of $141,583 thousand ^[27] and 112 stores in fiscal 2024 for $330,032 thousand ^[28], and Casey's states its store comparisons and notes in thousands ^[29], ^[30]. Table 5 divides the payments by the stores and prints $3,539,575 a store for fiscal 2026 and $2,946,714 for fiscal 2024. Those are chain purchases of stores that earn far more than this one, a ceiling the price here sits well under.

Decision. The study would treat the shortage as support for the idea and the 223 private stops as the field that caps the diesel price.

Demand: Gallons, Inside Sales and Who Buys Them

Demand in the model is built from gallons and sales per day, not households. In Table 8 the seller's last year sells 3,750 gallons a day, 912,500 gallons of gasoline and 456,250 of diesel a year. At stabilization the study's base case sells 4,250 a day, with the same gasoline and 638,750 gallons of diesel a year; the sponsor's pro forma sells 4,302.

The gasoline assumption is anchored. NACS's 2025 State of the Industry figures describe an average convenience store with 322 fueling transactions a day of 8.4 gallons each, roughly 2,705 gallons daily ^[31]. The seller's assumed 2,500 gallons of gasoline is about 92 percent of that store, and no lift in gasoline from the diesel expansion is assumed in any column.

Diesel is the assumption that matters. The base case adds an assumed 250 gallons a day for each of two added positions. The sponsor's pro forma uses the NACS ratio, the average store's daily gallons over the 9.8 fueling positions that stores selling fuel averaged ^[32], which is 276 a position. One state threshold frames the number: a Pennsylvania truck stop establishment for video gaming terminals must have sold, or be projected to sell, on average 50,000 gallons of diesel or biodiesel a month ^[33]. The seller's 38,021 gallons a month sits below the line and both stabilized columns above it, at 53,229 and 54,811 in Table 8. The format is the smallest that lifts the sponsor's volume over the line; no gaming income is modelled.

Inside sales follow the traffic. The base case assumes inside sales of $2,060,000 at stabilization, the seller's $1,670,000 lifted by truck-driver traffic. In-store sales across the industry came to $341.2 billion in 2025, per NACS ^[34], spread over 151,975 stores, 122,620 of them selling fuel, in the 2026 NACS count built on NIQ TDLinx data ^[35]. The assumed $2,060,000 is about 92 percent of the per-store average that implies, which the sponsor's pro forma uses in full at $2,245,106; the base case stays below it because the state's population is flat and the stop has no chain buying.

Foodservice carries the margin. NACS's 2025 data credit it with 28.5 percent of what the store sells inside but 38.9 percent of the gross margin earned there ^[36]. The model uses that share, with an assumed foodservice margin of 46.0 percent, 1.6 times the assumed 30.0 percent merchandise margin. NACS put fuel at 65.0 percent of 2025 sales dollars yet 38.8 percent of gross profit ^[37], within $817.5 billion of industry sales ^[38].

Decision. The study would build demand from gallons and inside sales per day and test the diesel numbers against truck counts at the site, which this example did not obtain.

The Site: Parcel, Tanks and Environmental Obligations

The site is described only by its role: an operating fuel stop adjacent to an interstate corridor in a rural-eligible area of Pennsylvania, with a store, eight fueling positions and a diesel tank of an assumed 20,000 gallons. No town, county or road is named, because the point is the arithmetic, not a location.

The expansion needs land, and the model buys 2.20 adjoining acres at an assumed $80,000 an acre, for $176,000: an assumed one acre for the island and truck circulation plus an assumed 0.06 acre a space. The size follows a state threshold. The Pennsylvania truck stop test for video gaming terminals requires at least 20 parking spaces for commercial vehicles, diesel islands, a convenience store, Lottery agent status and a parcel of a minimum acreage ^[39]. The format adds exactly those 20 spaces, so the spaces above the minimum are an assumed zero; a real study would confirm the combined parcel meets the test's acreage. A licensed establishment can run up to 5 video gaming terminals ^[40], but no gaming income is assumed here, and an entity is ineligible for a OneRD guarantee if more than 15 percent of its gross revenue, machine lease income included, comes from gambling, excluding State-authorized lottery proceeds ^[41].

Tanks are the largest regulatory exposure. A new diesel tank needs the double-wall containment and interstitial monitoring that 40 CFR 280.20 requires of systems put in after April 11, 2016 ^[42], and the assumed $330,000 diesel system cost, for a 30,000 gallon double-wall tank, piping, monitoring, electrical work and permits, is priced on that basis. Petroleum marketers must show financial responsibility of at least $1 million per occurrence ^[43], and owners of 1 to 100 tanks $1 million in annual aggregate ^[44].

Pennsylvania adds its own layer. USTIF coverage for an owner of 100 or fewer regulated tanks is $1.5 million per tank per occurrence and annual aggregate ^[45], and claims must be reported within 60 days after knowledge of a potential claim or a denial of coverage ^[46]. A lender would want a Phase I environmental assessment and the tank records before any commitment. The model carries an assumed $44,000 of existing insurance.

The collateral question is also site-specific. Under 7 CFR 5001.203, an appraisal's business or going concern value is taken off the reconciled market value, and the discount to collateral is applied to what is left ^[47]. The price of $1,020,000 buys land, building, fuel system, equipment and going concern together, so the lender's collateral is worth less than the price, and the loan of $1,810,000 exceeds the price by the expansion plus inventory and working capital.

Decision. The site is acceptable in concept and conditional on tank records, the acreage of the added land and the collateral split.

The Program Rules: USDA B&I Guarantee for an Acquisition and Expansion

A B&I guarantee backs a loan made by a private lender; it is not a direct USDA loan. The rules below come from 7 CFR part 5001 as the example's sources record it.

Eligibility. Eligible B&I projects include the purchase and development of land, buildings or infrastructure for commercial enterprises, including expansion, and business acquisitions, start-ups and expansions if jobs will be created or saved; an acquisition means buying an entire business ^[48]. Loan funds may buy equipment and fund working capital and inventory ^[49], and may pay the guarantee fee, lender fees, permit fees and feasibility studies ^[50].

Existing or new. An existing business has operated for at least a full year and reached full operational capacity or stable operations ^[51]; a new business has operated for less than a full year, or longer without reaching those conditions ^[52]. The buyer entity is new, but the business it buys has operated for years, and the classification matters for equity. The rule's table sets 10 percent for an existing business ^[53] and, for all other new businesses, 20 percent balance sheet equity or a borrower investment of 25 percent of total eligible project cost ^[54]. The sponsor's equity of $702,254 is 28.0 percent of cost against $251,225 for the existing-business test and $628,063 for the new-business borrower investment, both in Table 7, so it meets the rule however the Agency classifies the buyer. An existing business can also meet it with 10 percent balance sheet equity or debt to equity of no more than 9 to 1 at closing ^[55], and a new business with 20 percent or 4 to 1 ^[56]; the example tests the borrower-investment route only.

Feasibility study. The Agency may require a feasibility study when the lender's analysis cannot establish technical, market or economic feasibility, and for a guaranteed loan above $1,000,000 to a new business ^[57]. The model budgets an assumed $140,000 for third-party reports, closing and financing costs, which includes such a study.

Credit evaluation. The lender must substantiate projections that deviate from history, and increases in revenue, margins or profitability must be reasonable ^[58]. That is why the request rests on history. The lender must structure the debt for adequate coverage, supported by a cash flow analysis ^[59].

Terms. Final maturity cannot run past the useful life of the project or its collateral, nor past 40 years or a lower State limit ^[60]; guaranteed and unguaranteed portions share one amortization schedule with no balloon ^[61]. The rate is set by negotiation and may not be higher than the lender's usual rate on loans carrying no guarantee ^[62]. The Federal Reserve's H.15 release put prime at 7.00 percent for the first week of November 2026 ^[63]; the model adds an assumed 1.50 percent spread for a note rate of 8.50 percent, with an assumed 25-year amortization on the real estate share and 10 years on the inventory and working capital share.

Fees. For a loan under $5 million the FY 2026 OneRD notice sets an 85 percent guarantee, a one-time fee of 3.0 percent and a periodic retention fee of 0.55 percent, with a further 0.50 percent when the guarantee issues before construction is complete ^[64], effective October 1, 2025 ^[65]. The one-time fee multiplies the loan, the guarantee percentage and the fee rate ^[66]; the periodic fee is the retention rate times the December 31 balance times the guarantee percentage ^[67]. No later notice was found, so the example uses the FY 2026 rates. A reduced fee exists for listed criteria ^[68], such as long-term population decline or a persistent poverty county ^[69]; the example does not assume it. Neither the 90 percent maximum guarantee ^[70] nor the $25 million cap per borrower ^[71] constrains this project.

Coverage. Under the rule's definition, coverage is EBITDA less the capital spending on replacements that a borrower can foresee, divided by a year's principal and interest ^[72], so the model deducts a replacement reserve. The conditional commitment includes a minimum ratio only if the lender or the rule requires one ^[73], one of the items it addresses ^[74]. The only historical coverage test in the stored text is 1.1 times on the proposed debt service, and it applies where refinancing is a majority purpose of a loan to an existing business ^[75], ^[76]. This loan is not a refinancing. The B&I rule as cited therefore sets no new-loan or acquisition minimum, no SBA program floor applies, and the lender's assumed 1.25x minimum binds both the historical test and the stabilized test; Table 12 prints it as the binding minimum.

On gambling, the rule excludes State-authorized lottery proceeds from the count and treats lease income from machines as gambling revenue ^[77]; no gaming income is assumed.

Decision. A real study would set the request at the assumed lender minimum and test equity against both borrower-investment tests.

What the Purchase Costs and How It Is Funded: Uses and Sources of Funds

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

The inputs table is long because nearly every line is an assumption; the sourced ones are the prime rate, guarantee terms, equity tests, regional pump prices, NACS store averages, the broker's price range and the tank-capacity fee.

The price and its stated basis. A lender would want the price anchored first, and it is anchored here before any coverage is computed. CT Acquisitions, an M&A adviser, put 2026 convenience store acquisition multiples at 6x to 9x EBITDA for a multi-site portfolio with food, 4x to 7x for fuel plus a convenience store, and 3x to 5x for a single-site fuel-only business, as its own analysis ^[78]. This stop is fuel plus a store. Table 5 applies the multiple to the seller's EBITDA before the replacement reserve of $254,757: a multiple of 4.0, the low end of the range, and a price of $1,020,000 for the going concern with land, building, fuel system and equipment, rounded to $10,000. The seller's EBITDA after reserve is 22.8 percent of the price. CT Acquisitions says inventory typically runs $80,000 to $250,000 at cost and is always valued at cost ^[79], so the model buys an assumed $160,000 on top.

Three cautions belong with the price. First, the range is one adviser's own analysis, not a survey of closed sales. Second, the low end is a choice. The example prices an acquisition at the central value of the range, 5.5 times, and moves down only when equity would exceed the assumed 35 percent cap. In Table 6, the same smallest truck stop format at 5.5 times costs $2,892,764 and needs equity of 36.7 percent, over the assumed 35 percent cap, and the earlier format at the central price (four added positions, 50 spaces and an assumed 24 added staff-hours a day) costs $3,656,677 and needs 49.4 percent. Third, CT Acquisitions describes a total deal as business value plus real estate value ^[84] and says real estate is often separate or included for single sites priced at $500,000 to $4 million ^[85]. This example reads the CT Acquisitions range as covering a deal that includes the real estate; if the real estate were priced on top, the price would rise, equity would exceed the cap, and the project as set out here would not be the one financed.

The concepts tried. Table 6 shows the sponsor's figures only. The first format was dropped because about half its cost would be equity, the second was over the cap at the central price, and the third, the same format at the low-end price, is the first that fits. In every concept the loan the history supports is the smallest of the three amounts, so history, not projection, sets the request.

The uses side of Table 7 starts with the purchase of $1,020,000 and the assumed inventory of $160,000. The expansion follows: adjoining land of $176,000, the diesel island, tank, canopy and dispensers at $555,000, and truck parking improvements of $224,000. A construction contingency of $54,530 and interest during construction of $26,569 follow, then assumed third-party reports, closing and financing costs of $140,000 and an assumed working capital and expansion-period reserve of $110,000. Project cost before the guarantee fee is $2,466,099.

The guarantee fee is the one use that depends on the loan. On the requested $1,810,000 it is $46,155 (the loan times 85 percent times 3.0 percent), so the total project cost is $2,512,254, or $251,225 for each of ten fueling positions. If the guarantee issued before construction were complete, a memo line shows an added fee of $7,693 that is not in the total. The guaranteed portion is $1,538,500 and the lender holds $271,500 without a guarantee. The request is the smallest of three amounts: the loan the sponsor's stabilized pro forma supports at the assumed 1.25x, $2,999,061; the loan the seller's history supports, $1,819,525; and the assumed 80 percent of project cost before the fee, $1,972,879. It is rounded down to $1,810,000. Sponsor equity is $702,254, 28.0 percent of cost, under the assumed 35 percent cap, which is worth $879,289.

Equity at the supported loan. The request sits just under the loan the coverage supports, so the supported-loan figures differ little and belong together. The debt table prints a supported loan of $1,819,525, a guarantee fee recomputed on it of $46,398, a total project cost at that loan of $2,512,497, and equity at that loan of $692,972, which is 27.6 percent of cost. The request is $9,525 under the supported amount.

Decision. The study would present the price with its stated basis and its low-end position, show the central-price result beside it, and say that the equity fits the cap only because of that position.

The Pro Forma: Sponsor's Case, Study's Base Case and the Operating Statement

The revenue table, Table 8, sets three columns side by side: the seller's last year, the study's base case and the sponsor's pro forma. The seller's column is a record and is the same in both projected columns. The sponsor's pro forma is the model run at market figures; the study's base case replaces four of them with a site-specific estimate and gives a reason for each. Total revenue is $8,413,700 in the seller's year, $9,983,890 in the base case and $10,290,696 in the sponsor's pro forma.

Why the two projections differ. The sponsor's pro forma uses the 39.7 cents per gallon that OPIS data cited by NACS gave as the 2025 retail gasoline margin, against a five-year average of 38.3 cents ^[80]; the base case uses the seller's own realized margin of an assumed 36.0 cents, because 72 percent of drivers choose on price, a NACS survey result ^[22]. Inside sales are $2,245,106 against an assumed $2,060,000. Diesel per added position is 276 gallons a day against an assumed 250. And the sponsor's pro forma uses the NACS store-level wage of $15.04 ^[81] where the base case uses an assumed $16.00, the seller's own wage for a 24-hour store. The seller's year is identical in both, so the request, set by the historical test, does not depend on which projection is believed.

In Table 9 the base case has gasoline gross profit of $328,500, diesel $191,625, merchandise $441,870, foodservice $270,066 and other income of $45,000, for total gross profit of $1,277,061; the sponsor's pro forma has $1,380,490. Of the 20 to 25 cents a gallon NACS estimates it costs to sell fuel, about 8 cents is card fees ^[82], and retailers' card fees reached a record $21.3 billion in 2025 ^[83]. The model charges 8.0 cents a gallon on fuel and an assumed 2.0 percent on inside sales, for card fees of $165,300. Total operating expenses are $981,895, and EBITDA after the replacement reserve is $295,166, against $232,757 in the seller's year and $383,645 in the sponsor's pro forma.

Staffing in the expansion. The expansion adds staff in proportion to what it sells. The existing operation is assumed to use 48 labor hours a day, a 24-hour store with two on shift, plus a manager paid separately at an assumed $80,000. Added inside sales are staffed at the seller's own productivity, which lifts labor in Table 9 from 48.0 hours a day to 59.2 in the base case and 64.5 in the sponsor's pro forma. An earlier version of this example added a fixed block of hours unconnected to sales, and that single judgement absorbed nearly all the added gross profit. The productivity rule is why the expansion now adds income; it is a rule, not a measured schedule from any real stop.

An income comparable. Table 10 sets the Casey's average store, from the 10-K cited above, against this stop. Casey's average store open a full year sold 3,334 gallons a day at 42.57 cents of fuel revenue less cost of goods sold per gallon, and $2,199,000 of inside sales at a 40.7 percent margin, and earned average store operating income of $566,000 ^[25]. This stop sells more fuel in the base case, 4,250 gallons a day, but earns 33.53 cents a gallon blended because diesel is assumed at 30.0 cents, and its inside sales are an assumed $2,060,000 at a blended 34.6 percent margin. Its EBITDA of $295,166 is after a paid manager, professional services and advertising, against Casey's $566,000, a figure after store depreciation and before company overhead. The gap is the weaker single store, not inputs outside their anchors.

The ramp in Table 11 starts from the inherited level. In year 1 the island and parking are under construction, so the operation is the seller's: EBITDA of $232,757 and coverage of 1.26x. In year 2 an assumed 60 percent of the expansion volumes are realized, EBITDA is $250,845 and coverage 1.35x. In year 3 EBITDA is $295,166 and coverage 1.59x, on flat debt service of $185,231. Year 1 has no lift, so the historical test is the tight one.

What a lender would ask the seller for. Two years of tax returns and monthly point-of-sale reports, fuel invoices by product, the payroll register to check the assumed 48 labor hours a day, the real estate tax bill, the insurance schedule and the tank monitoring records.

Decision. The study would present the sponsor's case beside the base case, say that four inputs with stated reasons explain the difference, and size the request on history so that neither projection is needed to size it.

Coverage Under the Convention, Sensitivity and Break-even

The debt table, Table 12, prints a note rate of 8.50 percent, prime plus the assumed spread. The real estate share of the loan is $1,611,833 on 25 years and the inventory and working capital share is $198,167 on 10 years. Annual debt service is $155,747 on the first and $29,484 on the second, $185,231 in all, a loan constant of 10.23 percent. Against it, the base case EBITDA of $295,166 gives stabilized coverage of 1.59x, the sponsor's pro forma $383,645 gives 2.07x, and the seller's last year $232,757 gives 1.26x.

The minimum that binds. With no program floor in the B&I text as cited, the lender's own assumption of 1.25x binds, and the table prints it as the binding minimum in every column.

Two tests. A lender tests an acquisition on the seller's history as well as on projections. Historical coverage is 1.26x; stabilized coverage is 1.59x in the base case and 2.07x in the sponsor's pro forma. All three clear the assumed 1.25x. The historical test is the tighter and set the request: the seller's history supports $1,819,525 at the binding minimum, against $2,307,390 for the base case's stabilized earnings and $2,999,061 for the sponsor's. Cash flow above the minimum is $63,627 in the base case and $152,107 in the sponsor's pro forma.

Retention fee. With the first-year retention fee of $8,462 (the loan times 85 percent times 0.55 percent) added to debt service, stabilized coverage is 1.52x in the base case and 1.98x in the sponsor's. The base calculation leaves the fee out, which is another reason to call the pass thin.

Table 13 runs six changes against the base case. With revenue down 10 percent, EBITDA after the reserve falls to $183,990 and coverage to 0.99x; down 20 percent, EBITDA is $72,813 and coverage 0.39x. With the rate two points higher, annual debt service rises to $214,711 and coverage falls to 1.37x on stabilized earnings and 1.08x on the seller's history, under the assumed 1.25x. Operating expenses up 10 percent leave EBITDA at $196,976 and coverage at 1.06x. A diesel margin 5 cents lower leaves $263,228 and 1.42x. The last column keeps the expansion in place and removes its volumes altogether: revenue is $8,413,700, EBITDA $184,363 and coverage 1.00x, because the added fixed costs are carried with none of the added income. Only the diesel margin shock leaves both tests above the assumed 1.25x floor of the lender; the rate shock fails the history test, and a ten percent shock to volume or cost takes the stabilized case to or near 1.00x.

The break-even table, Table 14, asks what has to hold. Fixed operating costs are $816,595. Volume at 1.00x coverage is 90.1 percent of the base case and at the binding minimum 94.3 percent, which is 3,830 gallons a day at 1.00x and 4,007 at the binding minimum against 4,250. Inside sales would need to be $1,856,300 at 1.00x and $1,942,104 at the binding minimum, against the assumed $2,060,000. The diesel margin alone could fall to 12.8 cents at 1.00x and 20.0 cents at the binding minimum, on base volume, against the assumed 30.0 cents. The seller's history has no cushion beyond its 1.26x.

Decision. The loan the project can carry is $1,819,525 and the request is $1,810,000, so the structure clears, and the study would say plainly that the margin is thin on volume and cost.

What Would Change the Answer

A passing case needs no restructure, and none is offered. Five things would break it.

The price. The price of $1,020,000 sits at the low end of an adviser's range, and it is why equity is 28.0 percent. At the central 5.5 times, Table 6 puts the total project cost at $2,892,764 and equity at 36.7 percent, over the assumed 35 percent cap. If the real estate were priced on top, as CT Acquisitions allows for single sites ^[85], the price would rise further. Either reading breaks the structure before any operating figure moves, so the study would first ask for evidence that the seller will take the low-end price.

The seller's record. The request is set by the seller's history, which here is assumed, and the request of $1,810,000 sits only $9,525 under the loan that history supports at the assumed 1.25x. Verified earnings below the assumed record would put the request over the supported loan, so the study would replace the assumption with tax returns first.

Volume and cost. Break-even volume at the binding minimum is 94.3 percent of the base case, expenses up 10 percent take coverage to 1.06x, and revenue down 10 percent takes it to 0.99x. The assumed 250 gallons a day per added position, the assumed 30.0 cent diesel margin and the assumed inside sales of $2,060,000 each move the same line. Site traffic counts and supplier contracts that fix a margin would move these numbers, up or down.

The rate. The note floats on prime. A rise of 2 points leaves the stabilized case at 1.37x and takes the historical test to 1.08x, so a lender applying the assumed 1.25x to history would not lend the full request at that rate. The borrower would need a fixed rate or a smaller loan. A reduced guarantee fee ^[68] would trim a fee of $46,155 but would not touch the rate.

The gambling limit. No gaming income is assumed. A stop that reached the 50,000 gallons of diesel a month in Pennsylvania's truck stop test ^[33] and added video gaming terminals ^[40] would still be ineligible if its gambling revenue passed the OneRD ceiling ^[41], so the study would treat gaming as upside to be measured against that ceiling and not as base.

None of these is a recommendation to proceed. As set out, the project meets the assumed lender minimum on every test and has little room to spare on the tests that matter most.

FAQ: Choosing a Feasibility Study Company and Other Questions

Does the USDA B&I rule set a minimum debt service coverage ratio for this loan? The rule as cited names no figure for a new loan or an acquisition ^[73]; its stored 1.1 times historical test is confined to refinancing ^[75]. The lender's assumed 1.25x binds, and the base case clears it at 1.59x stabilized and 1.26x on the seller's history.

Why does the sponsor's pro forma show more coverage than the study's base case? The pro forma uses market averages for four inputs and the base case a site-specific estimate for each, with a reason. The pro forma shows 2.07x and the base case 1.59x; the seller's year, 1.26x in both, sizes the request.

How would a buyer choose a feasibility study company or consultant? For a guaranteed loan above $1,000,000 to a new business the Agency may require a feasibility study ^[57], and a lender will want the consultant independent of the seller and the broker. A buyer would ask a prospective company how it separates inherited earnings from projected ones, whether it will print the sponsor's pro forma beside its own base case and explain every difference, whether it will anchor the price to a stated basis before computing coverage, and whether it will say what would break a passing result. A company that shows only the stabilized year, or only the figure that passes, is a poor fit. The buyer would confirm with the lender, before engaging anyone, that the Agency will accept the consultant.