Genset Third Party Check: Records for Site Acceptance

A genset third party check is an independent witness test; this guide shows which records to verify, so you can compare inspection offers and request a scoped quote.

A genset third party check is an independent witness test that confirms a diesel generator set meets its agreed rating before a buyer accepts shipment. This guide sets out the documents an inspector needs, the load steps that get witnessed, and the measurements that belong in the signed record, so you can compare inspection offers on equal terms and request a scoped quote with confidence. Tesla Power builds open-frame and containerised sets for export projects, and the same acceptance run is used at the works and again at site handover. The practical answer is that an independent check does not replace the builder's own factory test. It adds a second pair of eyes to the numbers, and it confirms the report belongs to your serial number. A buyer who reads that file before payment can spot a thin margin while the set is still in the works, which is the cheapest moment to fix it. Ask for the as-found curve rather than a pass line. Confirm the instrument list and the calibration dates beside every reading. Get those inputs right and the rest of the purchase is straightforward.

  • Confirm the witness signed the as-found load curve, not just a pass line.
  • Ask for the instrument list and calibration dates with the record.
  • See our site planning guide for the room checks the factory run does not cover.
  • Keep the serial number on the report tied to the machine you receive.

Last updated 2026-11-19.

What does a genset third party check actually confirm?

Open-frame diesel generator set on the works floor with control panel and radiator visible before witness testing
Fig 1. The open-frame layout shows the engine, alternator and panel that the witness checks before the set is crated for export.

A genset third party check confirms three things at once, and the inspector signs to each one. It confirms the alternator holds voltage through the load steps, the engine carries that load without overheating, and the protection trips at the values printed on the schedule. None of those can be proved by a no-load run at the works gate. The witness also checks that the nameplate matches the order. The serial number on the report must be the machine being shipped. A generic certificate with a blank serial is not evidence you can rely on.

The check adds independence to the numbers. A builder testing its own set has an interest in a clean result; a hired inspector does not. That difference matters when two offers look identical on paper. Tesla Power Equipment supplies the same load ladder to the buyer and the witness, so both read from one schedule. The report then records the as-found curve, including the small wobble of real measurement, rather than a tidy line drawn after the event. Ask which standard the run followed. Name it on every page of the file.

Which documents does the inspector need before the run?

The inspector needs the purchase specification, the approved data sheet and the wiring diagram before the engine starts. Those three documents set the rating, the voltage band and the protection settings the witness will check. A run performed against the wrong revision is wasted, because the numbers cannot be tied to the contract. Send the same pack to the buyer at the enquiry stage. A short covering note listing each document and its revision saves an hour of argument on test day.

Add the calibration certificates for every instrument in the record and the duty class agreed for the set. A power analyser with an expired certificate gives a figure you cannot defend later. The duty class decides how long the full-load hold should last, so a standby set and a prime set are not witnessed to the same schedule. Tesla Power keeps a copy of the agreed schedule in the project file. List each document in a covering note and date it. Our load step acceptance note explains how the hold time is chosen for each class.

How do the load steps get witnessed on site?

The witness watches the load ladder climb in bands, and signs at each band. A 25% step comes first, then 50%, 75% and full rating. The engine must hold each band until coolant and oil temperatures settle. A short proof step above rating follows, held for a defined period. The inspector reads voltage and frequency at the alternator terminals and again at the distribution panel. A difference between the two points points to cable voltage drop on the test leads.

Every reading carries a time, an instrument number and the ambient conditions. That detail turns a trace into evidence. A step that recovers quickly is healthy. A slow sag that does not recover points to a governor or a fuel restriction, and the witness should log it even if the set still passes. The ISO 8528-1 steady-state definitions set the band, and the witness records the as-found value against it.

What does 500kVA deliver at 0.8 power factor?

The arithmetic is simple and checkable. Take the alternator rating of 500kVA and multiply by the power factor of 0.8. The result is 500kVA x 0.8 = 400kW of real power at 0.8 pf. At 400 V three phase, that real power corresponds to a line current of about 577 A. Hold that load for 2 h and the recorded energy is 400kW x 2 h = 800kWh. A witness who only writes the nameplate total misses both figures. Ask for the measured current beside the voltage at every step.

This difference matters at the quotation stage. A buyer who sizes a set from a kW figure alone may order a machine too small for a motor-heavy load. A buyer who sizes from kVA alone may pay for capacity the site cannot use. State the power factor in the enquiry, and check the same sums against our frames listed by rating before you freeze the specification. Tesla Power quotes both figures on every offer.

Which measurements belong in the witness record?

The record is the document a warranty claim will lean on. It should show the ambient conditions, the fuel grade, the oil grade and the instrument serial numbers. Every reading needs a timestamp and the name of the person who took it. The table below keeps the bands as percentages, so the same ladder works for any frame size. ISO 8528-5 describes the structure of the run and the order of the bands. Read the standard numbers together: the rating, the run structure and the measured noise describe one machine.

MeasurementAcceptance limitRecorded at
VoltageWithin 2.5% of 400 VEvery load step
FrequencyWithin 1% of 50 HzEvery load step
Oil pressureWithin the maker bandStart and finish
Coolant temperatureBelow the alarm pointSteady state
Load bank cables connected to the alternator terminals of a generator set during a staged third party witness test
Fig 2. Staged load steps prove the set holds voltage and frequency at each band before an independent signature is added.

Keep the raw analyser files as well as the summary table. Modern power analysers export a waveform capture, and that capture settles an argument about a single odd reading. Store the files with the generator serial number in the filename. A shared folder beats an email attachment, because the folder survives staff changes. Yangzhou Tesla Power keeps the same folder structure for every project, so the witness can find the file years later.

Why compare an independent report with the builder's file?

Comparing the two files is the cheapest quality control a buyer can do. A pass line that says OK tells you nothing about how close the set came to the limit. The as-found curve shows the real margin. A thin margin at full load is a warning you want before money moves. Ask the witness to sign the as-found data, and compare it with the rating on the nameplate before you confirm the order. A curve drawn by hand rather than measured is the first thing to question.

The other red flag is a report that cannot be tied to the actual set. The serial number on the report must match the nameplate and the packing list, and the date should sit before dispatch. If a supplier offers a generic report and promises to fill the serial later, that is not a test you can rely on. Ask for the as-found data tied to your machine. The works and the assembly bay are where that serial is applied, so the two documents should agree.

Common gaps in third party inspection reports

Reading only the headline pass mark is the first mistake. A pass at full load with a one-percent voltage margin is far tighter than a pass with a four-percent margin, yet both say pass. Read the margin, not the verdict. The margin is what survives a hot afternoon on site. The curve tells you the truth the stamp hides. Tesla Power flags this at the enquiry stage so it is not lost in a busy project.

A second mistake is confusing the no-load figure with the loaded figure. The set looks good at no load, and some reports lead with that number. Insist on the full-load band reading for voltage, frequency and noise. A third mistake is buying on a single power figure without stating the power factor, so a buyer compares a kVA offer with a kW offer and assumes they are the same. The duty class comparison note sets out which duty class belongs behind each figure.

Comparing inspection offers before you order

Put every offer on the same basis before you choose. Ask each bidder to state the rating class, the ambient condition it applies to, the alternator reactance, the fuel grade and the report format. An offer that leaves any of these blank is not cheaper; it is incomplete. A bidder who answers with a data sheet and a load step trace is giving you something you can check. Use the same question list for each bidder, so the answers can be compared line by line. Our compact export frame shows the level of detail a builder should supply.

Then check the terms against the technical answer. Compare the offers from genset manufacturers on one page, and ask Chinese genset suppliers for the dip figure in writing. A genset for export orders needs a graded enclosure checked against the site limit, and a genset for standby duty follows a shorter hold than a prime set. A genset for data centre projects adds a longer full-load run. The NFPA 110 weekly check sets the routine interval once the set is live.

Generator control cabinet with labelled buttons and a digital display beside a diesel set under independent inspection
Fig 3. A controller display only proves the set ran; the signed witness record is what a buyer can audit after handover.

Compliance is a separate file, and it should travel with the witness record rather than replace it. Ask for the genset CE marking requirements that apply to the destination market, and request the conformity assessment for the EU market alongside the load trace. These describe requirements, not claims. Tesla Power keeps the two files together so the buyer sees one consistent set of numbers.

Frequently asked questions about genset third party checks

A witnessed run for a 500kVA set takes roughly one working day, including the two-hour full-load hold and the proof step. Smaller sets finish faster, while parallel or enclosed packages need extra time for the enclosure noise check. Book the inspector before production starts so the slot is not the critical path. Confirm which bands will be shown and ask for the proof step to be included.

No. A factory test is run by the builder and a third party check is witnessed by an independent party. The builder's report proves the set ran; the witnessed report adds a signature that the numbers were read as stated. Ask for both documents, because a builder-only file leaves the independence question unanswered. The two are not interchangeable at handover.

Accept a steady-state voltage band of ±5% at full load, with the transient dip shown as well. A wider band means the set is outside its rating and may upset sensitive switchboard loads. Read the as-found curve at the full-load step, not the no-load headline, because the loaded value is what your site experiences each day.

Yes, provided the set has not been reconfigured after the factory run. A swapped enclosure or a changed protection setting invalidates the earlier figures. The site acceptance repeats the key bands on the delivered machine, and the serial number must match the factory file. Book the visit before commissioning so the two records can be compared side by side.

No, provided the run follows the manufacturer's procedure and the record is complete. Read the engine manual before the test. Some makers set a limit on how long the set may run at overload, and exceeding that limit is what causes a dispute. Keep the witnessed file with the warranty documents from the first day.

The run structure follows ISO 8528-5, the ratings and tolerances follow ISO 8528-1, and the noise measurement follows ISO 8528-10. For emergency supply, NFPA 110 adds the routine test intervals. Name the standard on every page of the report so the figures can be checked against a common reference.

A genset third party check is defined by the load ladder, the tolerance bands and the signed independent record. Size for the proof step, prove the voltage dip with a load run, and put every offer on the same basis. Request the witnessed file with the quote, and the choice becomes a checkable decision. To review a report against your load list, the Tesla Power team can work from your survey form.

Tesla Power Engineering Team — The engineering team specifies and commissions diesel generator sets from 10kVA to 2000kVA for export projects, with a focus on factory acceptance tests and inspection records.