Food-processing refrigeration efficiency feasibility: Oregon USDA Feasibility Study Case Study

    Last updated 2026-11-08

    How to read this case study. The engagement summary reflects the published record. Client names, actual costs, loan terms and proprietary data are confidential, so the project budget, market sizing and financial projections below are a modeled example built from public data for this project type and location. They show the analysis a lender-grade study performs, not the client's figures.

    Data current as of November 2026.

    Engagement summary

    This engagement assessed refrigeration energy-efficiency upgrades at a food-processing operation in Hermiston, Oregon. It was prepared in 2024 for a regional community bank, and its defining feature was energy-savings verification. That feature matters more now than when the study was written. Under the REAP rule effective 16 October 2026, a grant application must show 12 months of actual savings against 12 months of pre-installation data [1]. USDA also rescinded the FY2025 to FY2027 REAP funding notice in April 2026 [2]. This page describes the engagement type and models an Umatilla County food plant's refrigeration upgrade. It then shows how such a project is financed and verified under the rule in force.

    ItemPublished record
    ProgramUSDA Rural Energy for America Program (REAP)
    Asset classFood manufacturing
    ProjectFood-processing refrigeration efficiency
    LocationHermiston, Oregon (Umatilla County)
    RegionPacific Northwest
    Loan size tierUnder $5M
    Year prepared2024
    Lender typeRegional community bank
    OutcomePrepared with energy-savings verification

    Sources: the site's engagement index [12]; county from Census place files [6].

    The record describes the method. An energy audit established pre-upgrade consumption, verified that savings were measurable and fixed the eligible project cost. A post-installation monitoring plan then met USDA's savings-verification standard. The record also lists the grant cap, cost share and guarantee terms of the notice in force at the time. Those terms were superseded by the April 2026 rescission and the 2026 rule, and none of them is used here.

    The question the lender and USDA needed answered

    Refrigeration upgrades in a food plant are among the most reliable energy projects a rural lender sees. The compressors and fans run most of the year, and the measures are proven. The questions are about proof and timing:

    1. How much will the plant actually save? The audit has to split refrigeration from the rest of the plant's load and estimate each measure's saving. It also has to allow for measures overlapping, since a fan saving shrinks once the compressors run more efficiently.
    2. How will anyone know? A food plant's energy use rises and falls with production and weather. A raw comparison of bills before and after can show savings that are not there, or hide savings that are. The verification plan must normalise for both.
    3. Do the savings carry the loan? Unlike most solar, a well-chosen efficiency package can cover its own debt service. The study shows whether it does, and how fast.
    4. Can it be installed without disrupting production? Work has to fit the processing calendar.

    Under 7 CFR part 5001, the lender evaluates the project's credit and any feasibility study in writing [3]. Since 16 October 2026, a REAP grant for an efficiency project above $80,000 needs an energy audit, and projects of $200,000 or more need two years of pro forma statements and an engineer's completion certificate [1]. The rule also makes simple payback an eligibility test: it must not exceed the useful life [1].

    What changed for this engagement type on 16 October 2026

    TopicRule effective 16 October 2026 [1]What it means for a refrigeration upgrade
    TimingProject Period must end 12 to 24 months before the grant applicationThe loan builds the project; a grant follows a year of verified savings
    Data12 months of actual savings plus 12 months of pre-installation dataThe M&V plan becomes the grant application's evidence
    Award sizeAt most 25% of cost and $250,000 for energy efficiencyOn a $1.17 million project the $250,000 cap binds
    Combined fundingGrant plus REAP guaranteed loan at most 75% of project costA 75% loan leaves no room for a grant
    Loan aloneA REAP guaranteed loan without a grant is capped at 75% of eligible costs, with at least 25% from the borrower (7 CFR 5001.406(d))The plant funds a quarter even without a grant [3]
    EconomicsSimple payback must not exceed useful lifeAn easy test for this package; each measure is still checked
    LocationsProjects at more than one location ineligibleOne plant per application
    ApplicantOne application per federal fiscal year from each highest-level ownerA multi-plant processor chooses which plant goes first
    GamblingAny gambling income bars a grant; the loan keeps the 15% test in 5001.127(b)A revenue screen of the company and affiliates

    Project description (modelled)

    The modelled plant is an existing vegetable and potato freezing operation near Hermiston with an ammonia refrigeration system. It runs screw compressors, evaporative condensers, a spiral freezer and frozen storage. It uses 14 million kWh of electricity a year, 62% of it for refrigeration, and 180,000 therms of natural gas for sanitation hot water. It freezes 160,000 tons of finished product a year. The project is a package of six measures that the energy audit ranks by payback, installed over about four months in the off-season.

    Baseline

    Baseline input (model)ValueHow the energy audit sets it
    Plant electricity use14,000,000 kWh a yearTwelve or more months of bills and interval data
    Refrigeration share62% (8,680,000 kWh)Submetering or compressor logs
    Lighting in refrigerated areas700,000 kWhFixture inventory and hours
    Natural gas for sanitation hot water180,000 thermsGas bills; boiler logs
    Finished product frozen160,000 tons a yearProduction records
    Refrigeration energy per ton54.2 kWhComputed; the M&V normaliser
    Avoided electricity value$0.068 per kWhAssumption; the cooperative or utility tariff
    Demand charge / share of saved demand at the monthly peak$9 per kW-month / 50%Assumption; interval data
    Natural gas price$0.85 per thermAssumption; gas bills

    The efficiency measures

    Measure (model)Year-1 savingsYear-1 valueInstalled costSimple paybackUseful life assumed
    Compressor sequencing controls and a variable-speed trim compressor607,600 kWh$46,785$185,0004.0 years15 years
    Floating head pressure with variable-speed condenser fans484,344 kWh$37,294$140,0003.8 years15 years
    Variable-speed evaporator fans and controls303,522 kWh$23,371$165,0007.1 years15 years
    High-speed doors at freezer and dock openings145,691 kWh$11,218$120,00010.7 years12 years
    Desuperheater heat recovery to sanitation hot water39,600 therms$33,660$210,0006.2 years20 years
    LED lighting with controls in refrigerated areas420,000 kWh$32,340$160,0004.9 years15 years
    All measures (refrigeration savings applied in sequence)1,961,157 kWh and 39,600 therms$184,669$980,0005.3 years

    The package saves 1,961,157 kWh a year, 14.0% of plant use, and cuts refrigeration energy by 17.8%. Heat recovery saves 39,600 therms a year. At full realisation the year-one value is $184,669 against an installed cost of $980,000, a 5.3-year payback before soft costs. The refrigeration measures are applied in sequence so that the package does not count the same kilowatt-hour twice. The compressor controls take their share first, and each later measure saves a share of what remains. The high-speed doors are the weakest measure, at 10.7 years against a 12-year life. A sponsor could drop them to strengthen the package, or keep them for product quality and dock safety.

    The engagement record cites Energy Trust of Oregon incentives for refrigeration controls, variable-speed drives, electronic expansion valves and high-speed doors, and an Oregon Department of Energy technical-assistance program for rural food processors. Utility efficiency programs in Oregon depend on which utility serves the plant. The Oregon hub names Umatilla Electric Cooperative among the state's rural cooperatives [13]. The model counts no incentive until a program confirms the plant's eligibility. Both record claims are in VERIFY.

    What each measure does in a freezing plant

    Compressor sequencing and a variable-speed trim compressor. Large screw compressors are least efficient at part load, and a freezing plant spends much of the year at part load: between shifts, in cool weather and outside the peak harvest weeks. Sequencing controls run the fewest machines at their best loading and let a single variable-speed machine follow the swings. The audit sets the saving from compressor logs of suction pressure, slide-valve position and motor current. Nameplate ratings do not show it.

    Floating head pressure with variable-speed condenser fans. Many plants hold condensing pressure high all year as a precaution. Letting it fall with the outdoor wet-bulb temperature cuts compressor work in the cooler months, which in eastern Oregon is much of the year. The engineer sets the minimum pressure the expansion devices and hot-gas defrost still need, and the study's saving follows that limit, not a theoretical one.

    Variable-speed evaporator fans. Fans in freezers and cold rooms often run at full speed whatever the load, and every watt they use becomes heat the compressors must remove. Slowing them when the space is at temperature saves fan energy and compressor energy together. The audit's challenge is to count that interaction once, which is why the package applies savings in sequence.

    High-speed doors. Strip curtains and slow doors let warm, moist air into freezers. That adds load and frost, and frost means more defrost energy. The saving is real but smaller than the controls measures, and its value partly lies outside energy, in less ice on floors and coils.

    Heat recovery. A desuperheater takes the hottest part of the compressor discharge and uses it to preheat sanitation water. The plant would otherwise heat that water with natural gas. The study checks that the recovered heat matches when and how much hot water the sanitation shifts use.

    LED lighting with controls. In refrigerated space, a lighting watt saved is also a watt of heat the refrigeration system no longer removes. The model counts only the lighting saving, which keeps it conservative.

    Energy-savings verification

    The record's outcome, energy-savings verification, is now the core of a REAP grant file. A verification plan defines the baseline period, the meters, the variables that drive energy use and the method for adjusting the baseline when those variables change. For a freezing plant the main driver is tons of product frozen, with outdoor wet-bulb temperature second, because it sets condensing pressure. The example below shows why normalising for production matters.

    Reporting-year caseTons frozenMetered refrigeration kWhSavings by raw bill comparisonSavings normalised to productionError in the raw comparison
    Production as in the baseline year160,0007,138,8431,541,1571,541,1570%
    Production 10% higher in the reporting year176,0007,852,727827,2731,695,273-51%
    Production 10% lower144,0006,424,9592,255,0411,387,04163%

    If the plant freezes 10% more product in the reporting year, a raw comparison of bills understates savings by 51%. If it freezes 10% less, the raw comparison overstates them by 63%. Normalising to refrigeration energy per ton removes both errors: the plant moves from 54.2 to 44.6 kWh per ton frozen. A real plan uses monthly or weekly regressions against production and weather, submeters the compressor room and keeps the raw data for the reviewer. That evidence lets an application show "12 months of actual savings" that a reviewer can trust.

    Project cost and sources and uses (modelled)

    No construction-cost database sits behind these figures. Each line is a labelled assumption, and a real study replaces it with refrigeration contractor bids, vendor quotes for drives and doors, the auditor's proposal and the controls integrator's scope.

    Line itemModel amountSource in a real study
    Compressor sequencing controls and a variable-speed trim compressor$185,000Contractor and vendor bids
    Floating head pressure with variable-speed condenser fans$140,000Contractor and vendor bids
    Variable-speed evaporator fans and controls$165,000Contractor and vendor bids
    High-speed doors at freezer and dock openings$120,000Contractor and vendor bids
    Desuperheater heat recovery to sanitation hot water$210,000Contractor and vendor bids
    LED lighting with controls in refrigerated areas$160,000Contractor and vendor bids
    Controls integration, commissioning and M&V metering$75,000Controls contractor
    Contingency (8% of installed cost)$84,400Engineer judgment
    Energy audit (required for an efficiency grant above $80,000)$18,000Auditor proposal
    Feasibility study (within the cost page band for REAP loans above $600,000)$8,500Fixed-fee proposal
    Loan closing and legal$6,500Lender estimate
    Total eligible project cost (model)$1,172,400
    Cost outside the eligible budgetModel amountNote
    REAP guarantee fee (1.0% of the guaranteed portion)$7,034FY2026 REAP rate [4]
    Interest during a four-month installation (half drawn on average)$10,625Installed in the off-season
    Annual retention fee, year 1 (0.25% of guaranteed principal)$1,759[4]
    Utility or state efficiency incentivesNot countedTreated as upside until a program confirms eligibility

    Two ways to finance it

    UsesAmountStructure A: REAP loan onlyAmountStructure B: smaller loan, later grantAmount
    Efficiency measures, integration and contingency$1,139,400REAP guaranteed loan from the community bank (75%)$879,300REAP guaranteed loan$629,300
    Energy audit$18,000Owner contribution (25%)$293,100Owner cash at construction$543,100
    Feasibility study, closing and legal$15,000REAP grant$0REAP efficiency grant after 12 months of verified savings$250,000
    Owner cash returned if the grant is awarded-$250,000
    Total eligible project cost$1,172,400Total$1,172,400Total net of grant, if awarded$1,172,400
    Grant test (rule effective 16 October 2026)ModelSource
    25% of project cost$293,100[1]
    Energy efficiency maximum$250,000[1]
    Largest possible grant$250,000Lower of the two
    Room left under the 75% combined cap if the loan is at 75%$0[1]
    Structure B: loan plus grant$629,300 + $250,000 = 75.0%[1][3]
    Energy audit requiredYes: efficiency project above $80,000[1]
    Data for the application12 months of verified savings and 12 months of pre-installation data[1]

    The efficiency cap of $250,000 binds before the 25% limit ($293,100). Structure B therefore borrows $629,300 and asks the plant for $543,100 at construction, returning $250,000 if the grant is awarded after a year of verified savings. The grant is competitive, decided in one national competition for requests over $20,000 [1], and as of 7 November 2026 no funding notice under the rule had appeared. The community bank should underwrite Structure B on its own debt service. For a plant with strong cash flow, Structure B is a reasonable bet. For one with thin liquidity, Structure A is safer.

    Market analysis

    The market for an efficiency project is the plant's own future: whether it will run long enough, and at enough volume, to earn the savings. The study therefore looks at the regional processing base, the plant's raw-product supply and customer contracts, and the energy prices that set the value of each kilowatt-hour saved.

    The site and the rural test

    MeasureValueSource
    Hermiston city population, April 2020 estimates base19,422Census Vintage 2025 subcounty estimates [6]
    Hermiston city population, July 2025 estimate19,906[6]
    Umatilla County population, July 2025 estimate81,119Census Vintage 2025 county estimates [6]
    Umatilla and Morrow counties, OMB status (2023)Both non-metropolitanOMB list1_2023 [7]
    Oregon places above 50,000 in the April 2020 base12, none east of the Cascades except Bend[6]
    Nearest places above 50,000Kennewick (83,957), Pasco (77,324) and Richland (60,580), Washington[6]

    Hermiston had 19,422 residents in the Census April 2020 estimates base, below the 50,000 line in 7 CFR 5001.3. Umatilla County is non-metropolitan; Hermiston and Pendleton anchor a micropolitan area. The nearest places above 50,000 are Kennewick, Pasco and Richland across the state line in Washington. The rule's urbanized-area exclusion follows their contiguous urbanized area, which a dated USDA eligibility map determination will confirm for the parcel. The Oregon hub lists eastern Oregon, including Umatilla County, among the state's eligible regions [13].

    The Columbia Basin processing economy

    CountyPopulation, July 2025OMB status (2023)Manufacturers (employees), 2023Food manufacturers (employees), 2023Food manufacturing pay per employee
    Umatilla County, OR81,119Non-metropolitan61 (2,858)17 (1,356)$60,310
    Morrow County, OR12,602Non-metropolitan10 (1,466)6 (1,411)$58,656
    Union County, OR25,900Non-metropolitan27 (1,263)Not shownn/a
    Benton County, WA221,722Metropolitan152 (4,346)20 (1,354)$54,794
    Franklin County, WA102,612Metropolitan43 (3,604)16 (3,192)$54,537

    County Business Patterns 2023 (establishments with paid employees); OMB 2023; Vintage 2025 county estimates [5][6][7].

    Umatilla County's 17 food manufacturers employ 1,356 people at an average of $60,310. Neighbouring Morrow County's 6 employ 1,411, a sign of a few very large plants. Across the river, Franklin County's 16 food plants employ 3,192. The Columbia Basin is one processing region split by a state line, and a Hermiston plant competes for labour, raw product and freight with plants on both sides. For an efficiency study that is reassurance and a caution. The plant's market is deep, but its cost position against large neighbours is what the project defends.

    Oregon: food manufacturing (excluding meat), supply and rural footprint

    MeasureOregonUnited States
    Establishments, 2019 → 2023642 → 625 (-2.6%)24,589 → 27,414
    Per 100,000 residents (2023)14.78.1
    Average employees per establishment37.842.0
    Average annual pay per employee$54,991$58,248
    Share of establishments in non-metro counties16.1% (102 in 11 counties)
    Share of state population in non-metro counties (2025)15.0%

    Non-metro counties with the most food manufacturing (excluding meat) (2023): Clatsop County (15); Coos County (13); Lincoln County (13); Umatilla County (13); Hood River County (12); Tillamook County (11).

    Sources: U.S. Census Bureau, County Business Patterns 2019 and 2023 (state and county files; establishments with paid employees; employment and payroll suppressed in some cells); Vintage 2025 county population estimates; OMB metropolitan statistical area delineation, July 2023. Non-metro county is a proxy for USDA rural eligibility, not the eligibility test itself.

    StateFood plants 2019 (excluding meat)2023ChangePer 100,000 residentsEmployees per plantAverage payShare in non-metro counties
    Oregon642625-2.6%14.737.8$54,99116.1%
    Washington7608288.9%10.637.8$58,20313.6%
    Idaho19323119.7%11.775.3$57,84134.2%
    California3,5063,85510.0%9.839.4$60,4632.4%

    State data packs, County Business Patterns 2019 and 2023, food manufacturing excluding meat [5].

    Oregon has more food plants per resident than any of its neighbours (14.7 per 100,000), but it lost 17 plants between 2019 and 2023 while Washington, Idaho and California added plants. Consolidation and cost pressure are the backdrop. A plant that cuts 14% of its electricity and most of its hot-water gas is better placed to stay in the group that survives. Note one data point: the pack counts 13 non-metro Umatilla County food plants excluding meat, while the county file above counts 17 including meat.

    Financing benchmark: SBA lending to Oregon food manufacturers

    Oregon: SBA lending to food manufacturing (excluding meat)

    Period7(a) loans7(a) dollars504 loans504 dollars
    FY2010 to FY201670$26.5 million3$960,000
    FY2017 to FY201929$10.1 million5$4.4 million
    FY2020 to FY202337$25.0 million4$2.5 million
    FY2024 to FY202627$8.6 million2$2.9 million

    Average 7(a) loan, FY2020 to FY2026: $524,492. Average 504 loan: $904,167. Of the 70 7(a) loans approved in FY2010 to FY2016 for Oregon food manufacturing (excluding meat), 5.7% were charged off (national: 7.4% of 3,150), on average 42 months after approval.

    Source: SBA 7(a) and 504 FOIA loan data, as of 30 Sep 2026 (project state; disbursed loans only; cancelled and undisbursed loans excluded).

    Oregon food manufacturers have a better seasoned loss record than the nation (5.7% against 7.4%). The modelled REAP loan of $879,300 is larger than the average Oregon 7(a) food loan of $524,492, which is typical of an equipment package for an established plant.

    REAP activity in Oregon and its neighbours

    StateREAP grantsREAP grant dollarsAverage grantDollars per residentRank: dollars / per resident (of 53)REAP guaranteed loansB&I guaranteed loans
    Oregon106$21,311,544$201,052$5.0317 / 17114
    Washington156$16,076,249$103,053$2.0925 / 36014
    Idaho93$21,643,985$232,731$11.7716 / 7010
    California93$28,847,015$310,183$0.7311 / 50535
    Nevada10$2,780,525$278,053$0.9049 / 4804

    USAspending prime awards by place of performance, FY2024 to FY2026, for REAP (10.868) and B&I (10.768) [9]. These awards were made under the pre-2026 rules.

    Oregon took 106 REAP grants worth $21,311,544, averaging $201,052, and recorded one REAP guaranteed loan in the period. That one loan is a small precedent, but a precedent. Under the 2026 rule the guaranteed loan becomes the way to build an efficiency project, so the regional bank's familiarity with it matters. The Oregon hub reports that USDA revoked approved-lender status for ten OneRD lenders on 12 May 2026 [13]. A sponsor should confirm that its lender remains approved before the study is commissioned.

    Financial feasibility

    Five-year pro forma

    Line (Structure A)Year 1Year 2Year 3Year 4Year 5
    Savings realised while controls are tuned85%95%100%100%100%
    Electricity savings (energy and demand, tariff +3% a year)$128,358$147,762$160,206$165,012$169,962
    Natural gas savings (+2% a year)$28,611$32,617$35,020$35,720$36,435
    Total energy savings$156,969$180,379$195,225$200,732$206,397
    Added maintenance, M&V reporting and insurance-$23,165-$23,165-$23,165-$23,165-$23,165
    Net project benefit (project NOI)$133,804$157,214$172,060$177,567$183,232
    REAP loan debt service (7.25%, 10 years)-$123,877-$123,877-$123,877-$123,877-$123,877
    Retention fee-$1,759-$1,634-$1,501-$1,357-$1,203
    Project-level coverage1.07x1.25x1.37x1.42x1.46x

    The package covers its own debt service from year one, even with 85% of savings realised while the controls are tuned. Coverage rises to 1.25 times in year two and 1.46 times by year five as tariffs escalate. This is the key contrast with a solar project: a well-ranked efficiency package can be underwritten largely on its own savings.

    Global coverage

    YearPlant cash flow before project (model)Project NOIExisting debt service (model)New REAP debt service and feeCoverage beforeGlobal coverage after
    Year 1$3,200,000$133,804$1,600,000$125,6352.00x1.93x
    Year 2$3,200,000$157,214$1,600,000$125,5112.00x1.95x
    Year 3$3,200,000$172,060$1,600,000$125,3772.00x1.95x
    Year 4$3,200,000$177,567$1,600,000$125,2342.00x1.96x
    Year 5$3,200,000$183,232$1,600,000$125,0792.00x1.96x

    The plant's cash flow and existing debt service are modelled placeholders. On them, the loan moves global coverage from 2.00 times to 1.93 times in year one and back toward 1.96 times as savings grow. For a regional bank, the project is a modest add-on to a strong existing credit.

    Term choice

    Term at 7.25%Annual debt serviceProject coverage, year 1Project coverage, year 3Useful-life check (5001.402)
    7 years$160,5450.82x1.06xWithin the shortest measure life (12 years)
    10 years$123,8771.07x1.37xWithin the shortest measure life (12 years)
    15 years$96,3221.36x1.75xLonger than the door life; within controls and LED life

    A 10-year term sits inside the shortest measure life (the doors' 12 years) and covers the debt from year one. Seven years is too short for year-one coverage. Fifteen years would outlast the doors, so 5001.402's useful-life limit [3] points to a blended term or to dropping the doors from the package.

    Payback: the grant eligibility test

    Payback testValueSource
    Total project cost$1,172,400Budget
    Year-1 dollar value of energy saved at full realisation$184,669Measures table
    Simple payback6.3 yearsCost / annual value [3]
    Shortest measure life12 years (high-speed doors)Assumption
    Passes the grant eligibility test?Yes, with a wide marginPayback must not exceed useful life [1]

    Break-even and stress tests

    Break-even test (full realisation, today's prices)ValueModel baseMargin
    Share of modelled savings needed for 1.00x81%100%19%
    Share of modelled savings needed for 1.25x98%100%2%
    Electricity value needed for 1.00x$0.052 per kWh$0.068-24%
    Payback limit: savings share at which payback equals 12 years53%100%47%

    The project covers its debt if it delivers 81% of the modelled savings, and reaches 1.25 times only at 98%. That gap is why verification and commissioning matter. A package that delivers four-fifths of its promise still pays, but the margin above 1.25 times depends on the controls being tuned and kept tuned. The payback test is far from binding: payback would reach the doors' 12-year life only if savings fell to 53% of plan.

    Savings realisedElectricity at $0.055At $0.068 (base)At $0.085
    60%0.56x0.70x0.88x
    75%0.75x0.92x1.14x
    90%0.93x1.14x1.41x
    100%1.06x1.29x1.59x
    110%1.18x1.43x1.76x
    Fixed rate (10 years)Budget as modelled: coverage at full savings; owner 25%Overrun 10%Overrun 20%
    6.25%1.34x; owner $293,1001.22x; owner $322,4101.12x; owner $351,720
    7.25%1.29x; owner $293,1001.17x; owner $322,4101.07x; owner $351,720
    8.25%1.23x; owner $293,1001.12x; owner $322,4101.03x; owner $351,720
    9.25%1.18x; owner $293,1001.07x; owner $322,4100.98x; owner $351,720
    Delay caseBase year 1Installation slips into the processing season; half the measures deferred a year
    Project NOI, year 1$133,804$66,902
    Debt service and fee$125,635$125,635
    Project coverage1.07x0.53x
    Global coverage1.93x1.89x
    Grant clock (Structure B)12 months of savings after completionStarts only when the last measure is complete

    If installation slips into the processing season and half the measures are deferred a year, project coverage falls to 0.53 times in year one. Global coverage barely moves (1.89 times), but any grant is delayed, because the 12-month savings period starts only when the last measure is complete. Scheduling the work into the off-season, with equipment ordered early, is the mitigation.

    Equity and loan-size tests

    TestRequirementModel result (Structure A)Source
    REAP guaranteed loan size$5,000 to 75% of eligible project costs$879,300 = 75.0% of $1,172,4005001.406(d) [3]
    Borrower contributionAt least 25% of eligible project costs$293,100 = 25.0%5001.406(d) [3]
    Grant plus loanAt most 75% of Project CostStructure B: $879,300 = 75.0%4280.122(b)(20) [1]
    Studies, reports and audits inside the loanAt most 10% of the loan$26,500 = 3.0%Part 5001 [3]
    Equity certificationLender certifies at closing against a closing balance sheetPro forma closing balance sheet in the study5001.306(a)(2) [3]

    Technical and environmental feasibility

    • Refrigeration engineering. The audit reviews compressor curves, suction and condensing pressures, evaporator defrost practice and part-load operation, and confirms each measure's saving from logged data. Controls changes on an ammonia system are made by qualified refrigeration contractors, and the study confirms the contractor's experience.
    • Process safety. Ammonia systems are subject to process safety requirements. The study confirms that the changes fit the plant's process safety management documentation and that controls changes go through its management-of-change procedure. It names the regulators involved without assuming their findings.
    • Food safety and production. High-speed doors and evaporator controls must hold product temperatures. The study confirms that temperature logs will be kept through commissioning.
    • Heat recovery. The desuperheater must meet sanitation water temperatures without starving the condensers. The engineer sizes it against the hot-water demand profile.
    • Equipment. New, commercially available equipment. The 2026 rule's foreign-adversary bar applies to solar and wind components [1], so it does not reach refrigeration controls, though the study records origins anyway.
    • Environmental. Work inside an existing plant disturbs no ground. REAP grants carry environmental review requirements [1], and the study places them in the schedule.

    Management and governance feasibility

    The study confirms who owns the plant and its energy data, who runs the refrigeration system, and who is responsible for keeping the controls tuned after commissioning, which is the main driver of persistent savings. It names the M&V lead and the reporting cycle. It confirms that the applicant qualifies as a rural small business under SBA size standards, as the record notes was checked, and screens the company and affiliates under 5001.127 and, for a grant, under the 2026 rule's bar on any gambling income [3][1]. If the company owns several plants, the one-application-per-year rule means it must choose which plant applies first [1].

    Economic and community impact (modelled)

    Measure (modelled)ValueBasis
    Electricity saved a year at full realisation1,961,157 kWh (14.0% of plant use)Measures table
    Refrigeration energy saved17.8%Measures table
    Natural gas saved a year39,600 thermsHeat recovery
    Energy cost avoided over five years$939,702Pro forma
    Refrigeration energy per ton frozen, before and after54.2 to 44.6 kWhM&V normaliser
    Permanent jobs addedNone claimedProtects an existing employer's cost base

    The community case is a rural processor's lower cost base and a smaller load on a regional grid. The plant keeps its jobs and its raw-product purchases from local growers. REAP's 2026 scoring rewards energy saved per award dollar, the share of energy saved and environmental benefits [1], and the verified savings are the evidence for those points.

    Risks and mitigants

    RiskEffectMitigant the study documents
    Savings below planCoverage below 1.25x under about 98% realisationAudit from logged data; commissioning; savings tracked against the M&V baseline
    Controls drift after commissioningSavings fade over timeNamed owner of controls; alarms; annual retro-commissioning
    Production changes distort resultsSavings misreportedNormalised M&V by tons and weather
    Installation runs into the seasonHalf-year of savings lost; grant delayedOff-season schedule; early equipment orders
    Electricity price lower than assumedLower savings valueTariff sensitivity; demand savings shown separately
    Ammonia system safetyWork must fit process safety rulesQualified contractor; management of change
    Lender approval statusGuarantee unavailableConfirm the lender's OneRD approval before engagement
    Grant not awardedBridge not repaidStructure A as base case
    Plant closure or consolidationSavings lostGlobal credit review; supply and customer contracts
    Gambling or other ineligible income in an affiliateGrant barScreen all entities

    What the five Part 5001 components showed

    Component [10]What it tests hereWhat the modelled study shows
    EconomicBaseline, measures, cost against benefit$184,669 a year from $980,000 of measures; payback 6.3 years on total project cost
    MarketThe plant's future and energy pricesA non-metro processing region with deep competition; tariff tested from $0.055 to $0.085
    TechnicalEngineering, safety, commissioningProven measures applied in sequence; ammonia and food-safety checks
    FinancialCoverage, equity, sensitivityProject coverage 1.07x in year 1 rising to 1.46x; global 1.93x
    ManagementControls ownership, M&V, reportingNamed owner of savings persistence and verification

    Why efficiency studies are returned

    • Savings estimated from nameplate ratings instead of logged operation.
    • Measure savings added together instead of applied in sequence.
    • No verification plan, or one that compares raw bills with no production adjustment.
    • Energy prices taken from the average bill rather than the tariff's energy and demand components.
    • Incentives counted before a program confirms eligibility.
    • A loan above 75% of eligible cost, or a grant planned on top of a 75% loan.
    • An audit, study and engineer's report that disagree on cost or savings.

    Program terms that applied

    TermREAP guaranteed loan (FY2026)B&I guaranteed loan under $5 million (FY2026)Source
    Guarantee80%85%91 FR 11272 [4]
    Upfront guarantee fee1.0% of the guaranteed portion3.0%[4]
    Annual retention fee0.25%0.55%[4]
    Extra fee if the guarantee issues before construction is complete0.50%0.50%[4]
    Loan size$5,000 to 75% of eligible project costs; $25 million per borrowerUp to $25 million per borrower5001.406(c), (d) [3]
    Borrower moneyAt least 25% of eligible project costs10% for an existing business5001.406(d); 5001.105(d) [3]
    TermUseful life, at most 40 years, no balloonSame5001.402 [3]
    RateNegotiated; fixed, variable or bothSame5001.401 [3]
    Rural testOutside cities over 50,000 and their contiguous urbanized areasSame5001.3 [3]
    Gambling screenMore than 15% of gross revenue from gambling bars the loanSame5001.127(b) [3]
    Feasibility studyWhen the lender or Agency deems it necessaryRequired over $1,000,000 to a new business5001.306(a)(3), 5001.307 [3]

    These FY2026 rates apply to loans obligated in FY2026 unless a later appropriation changes them [4]. The Oregon hub also lists REDLG pass-through loans through rural electric cooperatives and state programs as possible companion sources [13]. The study should confirm any of these before counting them, and check how non-REAP funds count against the rule's cost-share and funding caps.

    Lessons for sponsors of similar projects

    1. Rank the measures. In the model, controls, condenser and fan drives, lighting and heat recovery pay back in roughly four to seven years, while the doors take nearly eleven. Build the package from the audit's ranking, not a vendor's catalogue.
    2. Apply savings in sequence. Overlapping measures cannot each claim their stand-alone saving. Sequenced savings are the honest number and the one a reviewer will check.
    3. Verify against production. Normalise to tons and weather. A raw bill comparison can halve or inflate the savings.
    4. Install in the off-season. The grant clock and the year-one savings both depend on finishing before the processing season.
    5. Keep the controls tuned. Coverage above 1.25 times depends on near-full realisation, and controls drift without an owner.
    6. Plan for 25% of your own money. The REAP loan stops at 75% of eligible cost, and the efficiency grant is capped at $250,000.
    7. Treat the grant as a later rebate. It comes after a year of verified savings, through a national competition.
    8. Confirm your lender. Check that the bank is an approved OneRD lender before commissioning the study.

    The site's cost page lists REAP studies at $4,900 to $8,500 over 4 to 5 weeks for loans under $600,000, and $7,500 to $11,500 over 5 to 7 weeks above it, each as a firm fixed fee in a written proposal [11]. For a food processor planning refrigeration, heat-recovery or lighting upgrades, an independent USDA feasibility study consultant can tie the audit, the verification plan and the lender's credit case together.

    Request a scoped proposal for your REAP efficiency feasibility study at the contact page.

    David Miller, CPA

    USDA Feasibility, Inc.

    Frequently Asked Questions

    Can a food plant finance refrigeration upgrades with REAP after October 2026?

    Yes. A REAP guaranteed loan under 7 CFR part 5001 can finance up to 75% of eligible project costs for an energy efficiency improvement, with at least 25% from the borrower and an 80% guarantee in FY2026. A REAP grant of up to 25% of cost, capped at $250,000 for efficiency projects, can follow only after the measures have run long enough to show 12 months of actual savings against 12 months of pre-installation data.

    Do I need an energy audit for a REAP efficiency project?

    For a grant, yes, when the efficiency project costs more than $80,000. Smaller projects can use an energy assessment. The audit sets the baseline, estimates each measure's saving from logged operation and supports the eligible project cost. For a REAP guaranteed loan, the lender and Agency rely on the audit and, where they deem it necessary, a feasibility study. In the modelled case the audit costs $18,000 inside an eligible budget of $1,172,400.

    What does energy-savings verification mean for a REAP application?

    A plan that proves the savings after installation. It sets a baseline period, installs meters on the refrigeration system and records production and weather. It then adjusts the baseline so that a busier or quieter year does not distort the result. Under the 2026 rule, the grant application must show 12 months of actual savings, so the verification plan produces the application's evidence. In the model, a 10% swing in production would distort a raw bill comparison by more than half.

    How much can refrigeration upgrades save in a food plant?

    It depends on the system and how it is run, which is why the audit uses logged data. In the modelled freezing plant, six measures cut refrigeration electricity by 17.8% and total plant electricity by 14.0%. Heat recovery saves 39,600 therms of gas. The package returns $184,669 a year on $980,000 of measures. Controls and drive measures paid back fastest; high-speed doors were the slowest.

    Will the savings cover the REAP loan payments?

    For a well-ranked efficiency package, often yes. In the model, a $879,300 loan over 10 years at 7.25% costs $125,635 a year with the retention fee. Savings cover it 1.07 times in year one while controls are tuned, and 1.46 times by year five. The project breaks even at 81% of modelled savings, but reaches 1.25 times only near full realisation, so commissioning and controls upkeep matter.

    Is the efficiency grant worth waiting for?

    It is worth planning for, but not depending on. The maximum is $250,000 or 25% of cost, whichever is lower, and it is paid only after a year of verified savings through a national competition. Planning for it means borrowing less at the start: in the model, $629,300 instead of $879,300, with the plant bridging $250,000 until the grant arrives. A plant with strong liquidity can take that bet; a thin one should not.

    Is Hermiston, Oregon eligible as a rural area?

    On the data here, yes. Hermiston had 19,422 residents in the Census April 2020 estimates base, below the 50,000 limit in 7 CFR 5001.3, and Umatilla County is non-metropolitan. The nearest places above 50,000 are Kennewick, Pasco and Richland in Washington, and the exclusion follows their contiguous urbanized area. The Oregon hub lists Umatilla County among eligible regions. Attach a dated USDA eligibility map determination for the plant's parcel.

    Can utility incentives be combined with REAP?

    Possibly, but count them only once a program confirms the plant's eligibility. Which incentives apply depends on the utility serving the plant, and the engagement record's references to particular Oregon programs were not verified for this page. Under the 2026 rule, REAP grant plus REAP loan may not exceed 75% of project cost. The study should also check how other funds count against REAP's cost-share rules before the sources and uses are final.

    How long should a REAP efficiency loan run?

    No longer than the useful life of what it finances, under 5001.402. In the model the shortest-lived measure, high-speed doors, lasts about 12 years. A 10-year term stays inside that life and covers the debt from year one, while 7 years would not. A 15-year term would outlast the doors, so it needs a blended structure or a package without them.

    How long does a REAP efficiency feasibility study take?

    This site's cost page lists 4 to 5 weeks for REAP studies under a $600,000 loan and 5 to 7 weeks above it, each quoted as a firm fixed fee in a written proposal. Most of the time goes into the audit data, interval meter records, production history and the verification design. Having twelve months of energy and production data ready at kickoff shortens the schedule.

    Sources

    [1] USDA RBCS, Unleashing American Energy and Economic Prosperity; Rural Energy for America Program (REAP), final rule with comment period, 91 FR 62600 (1 Oct 2026), effective 16 Oct 2026; sections 4280.120, 4280.122, 4280.125, 4280.131, 4280.140. https://www.govinfo.gov/content/pkg/FR-2026-10-01/html/2026-20178.htm (accessed 7 Nov 2026)

    [2] USDA RBCS, Notice of Rescission of Funding Opportunity for the Rural Energy for America Program, 91 FR 20090 (15 Apr 2026). https://www.govinfo.gov/content/pkg/FR-2026-04-15/html/2026-07332.htm (accessed 7 Nov 2026)

    [3] 7 CFR part 5001, OneRD Guaranteed Loans (CFR 2026 edition, title 7 vol. 15, revised 1 Jan 2026), sections 5001.3, 5001.105, 5001.106, 5001.127, 5001.202, 5001.306, 5001.307, 5001.401, 5001.402, 5001.406. https://www.govinfo.gov/content/pkg/CFR-2026-title7-vol15/xml/CFR-2026-title7-vol15-part5001.xml (accessed 8 Nov 2026)

    [4] USDA, OneRD Annual Notice of Guarantee Fee Rates, Periodic Retention Fee Rates, Loan Guarantee Percentage and Fee for Issuance of the Loan Note Guarantee Prior to Construction Completion for Fiscal Year 2026, 91 FR 11272 (9 Mar 2026). https://www.govinfo.gov/content/pkg/FR-2026-03-09/html/2026-04581.htm (accessed 6 Nov 2026)

    [5] U.S. Census Bureau, County Business Patterns 2019 and 2023, state and county files. https://www2.census.gov/programs-surveys/cbp/datasets/ (accessed 7 Nov 2026)

    [6] U.S. Census Bureau, Vintage 2025 population estimates, county and subcounty totals. https://www2.census.gov/programs-surveys/popest/datasets/2020-2025/ (accessed 8 Nov 2026)

    [7] OMB metropolitan and micropolitan statistical area delineation file, July 2023 (list1_2023). https://www2.census.gov/programs-surveys/metro-micro/geographies/reference-files/2023/delineation-files/list1_2023.xlsx (accessed 7 Nov 2026)

    [8] U.S. Small Business Administration, 7(a) and 504 FOIA loan data, as of 30 Sep 2026. https://data.sba.gov/dataset/7-a-504-foia (accessed 7 Nov 2026)

    [9] USAspending.gov, award counts and amounts by place of performance for Assistance Listings 10.768 (B&I) and 10.868 (REAP), 1 Oct 2023 to 30 Sep 2026, API v2. https://api.usaspending.gov/api/v2/search/spending_by_award_count/ (accessed 7 Nov 2026)

    [10] OneRD Guaranteed Loan Regulation, 85 FR 42494 (14 Jul 2020), appendix A to subpart D of part 5001, Elements of an Acceptable Feasibility Study. https://www.govinfo.gov/content/pkg/FR-2020-07-14/pdf/2020-13991.pdf (accessed 7 Nov 2026)

    [11] usda-feasibility-study.com, USDA Feasibility Study Cost & Timeline (site page). https://usda-feasibility-study.com/usda-feasibility-study-cost (read 7 Nov 2026)

    [12] usda-feasibility-study.com, USDA Feasibility Study Engagements (published engagement record). https://usda-feasibility-study.com/engagements (read 11 Nov 2026)

    [13] usda-feasibility-study.com, USDA Feasibility Study in Oregon (state hub). https://usda-feasibility-study.com/oregon (read 13 Nov 2026)

    Ready to Start Your USDA Project?

    Get your free consultation today and take the first step toward securing USDA funding for your rural development project.

    Our USDA feasibility study consultants scope every engagement against the five Part 5001 components in a free consultation.

    Get Your Free Consultation

    Fill out the form below and we'll contact you within 24 hours to discuss your project.

    We'll respond within 24 hours