Compare gasoline and diesel generator quotes by duty, usable output, fuel cost, service, market evidence, logistics and total cost.
A gasoline vs diesel generator decision has no universal winner. First reject or return any offer that cannot prove the required duty, electrical build, destination route, or site fit. Then compare the surviving quotes by useful output, fuel cost, service, delivered scope, and buyer-owned operating scenarios. Keep the decision on HOLD while a material input is missing.
Pass the hard comparability gate first
Do not compare fuel price or purchase price until both offers describe the same buyer task. A shared model number or similar headline output proves very little.
Honda's official EU7000iS page lists 5,500 W rated and 7,000 W maximum for one 120/240 V gasoline inverter model. Kubota's official GL7000 page lists 6.5 kW prime and 7.0 kW standby for one 120/240 V diesel model. Both names contain 7000, but their sustained ratings and allowed duties are not the same.
The photo below shows the type of visible detail a buyer may receive. It does not identify the fuel, model, power, duty, or market approval of the offered set.
Use the same hard gate for both quotes. A failed row does not earn a low score; it goes back to the responsible party for evidence or correction.
| Field | A quote can pass when | Keep it on HOLD when |
|---|---|---|
| Product identity. | The make, model, suffix, fuel, and build name one set. | The quote mixes a family name, several variants, or a sample photo. |
| Power basis. | Rated and short-term values have clear names and units. | Rated, maximum, prime, or standby has no use limit. |
| Electrical output. | kW and kVA are separate. Power factor is stated when needed. | kW is ranked against kVA, or voltage, frequency, or phase differs. |
| Output boundary. | Both values refer to the complete set at the same output point. | Engine power, alternator size, and net set output are mixed. |
| Load response. | The exact set has data for any load that is sensitive to voltage or frequency change. | A key load has a limit but the supplier gives no matching proof. |
| Supplied build. | Enclosure, outlets or terminals, controls, and options are clear. | One price includes destination parts that the other leaves out. |
| Document identity. | The quote links to current data sheets and manuals for that build. | Files cannot be tied to the quoted model and revision. |
The formula kW = kVA × power factor helps only when the supplier gives the right power factor. It cannot fix a missing duty rule, wrong phase, or different output boundary. Here, HOLD is a buyer control; it is not an ISO rating term. The rated-versus-peak guide explains power labels, while the voltage and frequency checklist covers the destination supply.
Match each exact set to the real duty
Fuel does not grant a prime, standby, or continuous rating. The exact maker's data must allow the planned work. The public ISO 8528-1:2018 scope confirms that set ratings and uses need defined terms, but it does not rate a quoted model.
Write a short duty sheet for each offer. Include:
- The role: emergency backup, planned prime power, base load, rental work, or another clear use.
- Hours per event, day, and year. Keep test-run hours separate.
- Running kW and kVA by time block, not one large total.
- Motor or compressor starts, their order, and the load's dip and recovery limits.
- Site height, air temperature, dust, humidity, airflow, and any other limit that affects output.
- Loads that must not stop and the proof needed for them.
Caterpillar's set-rating guide is one maker's example of ratings tied to time, load, and use. It does not define another brand's set. A gasoline manual also cannot prove that every diesel set can run prime power.
Each model must pass the duty on its own, using the supplier's controlled source for the exact offered build. If a maker gives no valid rating or site correction, record HOLD — DUTY NOT PROVEN. Do not turn that gap into a low score or a fuel-type assumption.
Build a same-basis fuel-cost model
Litres per hour alone cannot show which set costs less to run. Match the load point, useful kW, allowed duty, test basis, and dated local delivered fuel price. Keep supplier evidence and buyer assumptions in separate columns.
For each load bin i:
fuel cost_i = hours_i × fuel use_i (L/h) × delivered price_i (currency/L)
useful energy_i = hours_i × delivered power_i (kW)
fuel cost per useful kWh = total fuel cost ÷ total useful energy
Use only load points that the source supports. Do not draw a smooth curve from two points or scale fuel use in a straight line without maker data. Leave a load bin blank if no sound input exists.
Honda lists a 5.1 US gal tank, 6.4 hours at rated load, and 16.0 hours at quarter load for the cited EU7000iS. That shows why load matters; it is not a full fuel curve. Kubota's current GL7000 sheet lists load-point data, but its shown tank, use, and run-time figures do not all give the same result at the printed precision. Ask for the test basis and rounding method. Do not guess which field is right.
The buyer should date each local fuel price and apply delivery fees, taxes, and minimum-drop costs in the same way. Keep expected operating hours as a buyer-owned input. Do not insert a global fuel-efficiency claim. An inverter gasoline set and a fixed-speed diesel set may also act quite differently at part load.
Price maintenance and support for the exact build
Count tasks from the right manual. Do not judge service cost by the presence of one spark plug or fuel-system part.
For each task, record the manual and revision, hour or date limit, first-service rule, part code, fluid type and amount, labour time, and local source. Add a separate row when heat, dust, wet weather, storage, or heavy use changes the plan. If a system is not on the offered build, do not price it.
Support needs facts too. Name the service party, territory, parts stock, tools or software, response terms, exclusions, and remedy. Put any unpriced or unowned support gap beside the relevant scenario. Phrases such as “local support” or “24/7 help” have no cost value until they are tied to a firm scope.
Honda's EU manual index maps EU7000iS manuals to EEJD serial ranges. The AG/AN manual for EEJD 1382344–1399999 also calls for more service in some harsh uses. It gives no one rule for all gasoline sets or all sites.
Do not assume a fixed service life for either fuel. If a supplier cannot prove design life or overhaul terms for the exact model, use the same buyer-selected study period for both offers. Test any replacement value as a separate case.
Verify that fuel can stay ready at the site
Cheap fuel is of little use when the correct grade cannot reach the site or remain fit. This is a local plan, not a feature of the generator alone.
The buyer and local qualified parties should answer five questions:
- Which grade is required, and who can supply the needed volume?
- Can deliveries reach the site during the planned duty or an outage?
- Which tank or container, permit path, and site rules apply?
- How will the team control water, dirt, ageing, heat, and cold?
- Who owns refuelling, tests, spill response, and each approval?
US OSHA 1910.106 shows how detailed one market's liquid-fuel rules can be. It is not a global tank design. FEMA's emergency-power guide and Sandia's stored-fuel study discuss fuel ageing and care. They do not set one shelf life or test cycle for every project.
If grade, supply, storage, climate, or local approval is still open, fuel cost is not yet a final decision input. Mark the affected scenario HOLD and assign the local question to its actual owner.
Separate destination evidence from noise evidence
Fuel is not a market certificate. The route may change with the country, engine class, power, mobile or fixed use, model year, and relevant sale or install date. Site permits may add more limits.
For each offer, request the engine make, model and family; set model; labels; approval or declaration numbers; holder; issue and expiry dates; linked reports; and a way to check each record. Also ask which party owns import, site, and operating approval. Put a quote on hold when an engine paper is used as proof for the whole set without a valid link. Do not use a noise figure to close an emissions or market-access field.
Noise needs its own basis. Record whether the value is sound pressure or sound power, plus load, distance or work point, mode, enclosure, and test method. Honda's UK EU70iS page lists 75 dB(A) sound pressure at the work point and 91 dB(A) guaranteed sound power. Kubota's GL7000 sheet lists 66 dB(A) at full load and 7 m. These are different measures, so they cannot prove that one fuel is quieter.
The US EPA has separate pages for nonroad engines and stationary-engine compliance. The EU has scoped engine and outdoor-noise rules. The UK outdoor machinery noise guide also splits Great Britain and Northern Ireland. None of these pages approves an unnamed set for an unnamed market.
Compare the delivered build and site work
Fuel does not tell you whether a set is easy to ship, move, connect, or service. Compare the delivered build, not a bare generator price.
Ask for product and packed size, net and gross mass, package count, lifting or rolling parts, drawings, cable and terminal scope, outlets, controls, batteries, fluids at shipment, cooling and exhaust needs, and start-up work. The quote should state what is included and what remains for the buyer.
Trade terms need the rule version and named place. Record the handover point, freight scope, risk transfer, and insurance owner. A unit price or loose freight estimate is not a landed scope.
Honda lists 263.2 lb dry, folding handles, and a wheel kit for the current US EU7000iS variants shown on its page. An older EEJD manual lists 261.5 lb with battery. These facts apply to different records and cannot be swapped. They also do not prove a safe lift plan.
Use gross packed data for freight. Use the build drawing and manual for site work. A catalogue photo cannot replace either one.
Award by scenario only after the gates pass
There is no fixed gasoline-versus-diesel break-even point. First hold or reject any offer that fails a duty, electrical, market, safety, or site need. Then compare only the offers that remain, using the same currency, time basis, and scenario inputs.
Keep the cost model plain:
| Cost group | Include only when the basis is known |
|---|---|
| Landed purchase. | Destination-ready build, buyer-paid freight and insurance, non-recoverable duty or tax, local handling, install, compliance work, and start-up. |
| Yearly use. | Load-bin fuel, planned parts and labour, fuel care, firm service support, and a buyer-approved cost for lost time. |
| Later years. | Supported replacements and residual value, with one currency, date, study period, and real or nominal discount basis. |
The ICC Incoterms® 2020 overview helps assign trade cost and risk. It does not give a generator cost formula. NIST Handbook 135 explains present-value methods in a US federal setting, but its inputs are not the buyer's tax or finance terms.
Run at least three buyer-owned cases: expected use, low use or a short life, and high use or fuel-price stress. If the winner changes, state which input caused the change and where the decision boundary moved. Do not hide the switch behind one total.
For a comparable BEAR shortlist, first review the generator catalogue. Then send the quotation team your market, use, load sheet, voltage, frequency, phase, yearly hours, quantity, and any models under review. Ask for the exact build, source files, open-field owners, and scenario inputs in the return. A fuel recommendation should follow that controlled comparison, not precede it.
