ETO Sterilizer FAT: What the Factory Test Must Prove Before Shipment

Supplier Qualification · Factory Acceptance Test

Published by Hangzhou Bocon | ETO Sterilizer Manufacturing & Validation Support

A factory acceptance test is the last moment when a chamber can still be corrected without a crane, a customs broker and a site shutdown. It is also the step buyers resource most thinly: one engineer travels, watches the machine run, and signs a report nobody reads until an auditor asks for it. This guide sets out the ten physical tests that make up a credible ETO sterilizer FAT, what the report must contain, and how to review the data before shipment.

Why a FAT Decides More Than a Machine Demo

The FAT takes place at the manufacturer’s works, before packing and shipment. At that point the chamber, the door mechanism, the vacuum system, the vaporizer, the control cabinet and the recipe software are all present and accessible, and every utility the machine needs is on tap. A leak found here is a workshop task. The same leak found after installation is a crane operation, a re-insulation job, a delay to the site qualification schedule and, often, a re-run of temperature distribution work already paid for.

A FAT is not process validation. It does not use your product, your load configuration or your bioburden, and it cannot demonstrate a sterility assurance level. What it demonstrates is that the equipment was built as specified, that its subsystems perform to defined limits, that the control system executes the intended cycle, and that the records needed later can actually be produced. Performance qualification on a loaded chamber comes later, on site. The distinction matters because it defines what a buyer should refuse to accept as “already covered” by the FAT.

What the Protocol Must Contain Before Any Test Runs

The protocol is a signed document, not a test sheet. It should identify the exact unit by its unique works number rather than a model designation, because the report will be filed against that unit for its entire service life. A protocol that refers only to a model cannot later prove which chamber was tested.

Beyond identification, six elements determine whether the test results will survive an audit. They are listed below with the failure mode each one prevents. The pattern is consistent: weak protocols are weak in what they fix in advance, not in what they measure on the day.

The Ten Tests That Make Up Physical Qualification

BOCON’s standard FAT protocol is built from a trial operation plus ten physical tests, followed by a full-cycle evaluation. The ten tests fall into five families: pressure integrity, vacuum performance, thermal performance, humidity control and air movement. Each answers a question that will be asked again during installation and operational qualification, which is why the data has to be recorded rather than merely observed.

Pressure integrity: positive pressure and vacuum leak

Two leak tests are needed because the seal is loaded in two directions. Under positive pressure the gasket is pushed outward; under vacuum it is drawn inward. A door that passes one can fail the other, and the vacuum test is the more demanding of the two because any ingress path also admits moisture. Both tests are run on a sealed, empty chamber with the temperature stabilised above ambient, held long enough for the pressure decay to be distinguished from thermal drift, and reported as a rate rather than a start and end pressure.

Vacuum performance and venting

Vacuum speed records how long the system takes to reach defined vacuum levels, and it separates a healthy pump from one that is merely still running. The test also establishes the deepest vacuum the machine can hold, because that figure governs how effectively air can be removed before gas admission. Venting is its counterpart: the chamber must return to near-atmospheric pressure through a filtered path quickly enough not to extend the cycle, and slowly enough not to disturb the load or the filter.

Thermal performance: heating rate and temperature uniformity

Heating rate is measured from a stable starting temperature to the working set point, and it tells the buyer how much of the cycle is spent waiting rather than sterilising. Temperature uniformity is the more important of the two. It is measured with a documented sensor map, at positions chosen to represent the chamber volume rather than convenience, and it identifies the cold and hot positions that will later be monitored during performance qualification.

The number of sensors is not arbitrary. ISO 11135:2014 sets a minimum related to chamber volume, and the map has to be attached to the protocol so that a later auditor can see where each channel sat. A uniformity result quoted without a map is not evidence of anything; it is a number.

Humidity control: humidification and humidity uniformity

Humidity is the parameter most often taken for granted and most often responsible for a marginal biological indicator result. The humidification test measures how quickly the chamber moves from a dry state to the working humidity set point, and whether the pressure rise that accompanies steam injection is consistent with the humidity achieved. The uniformity test then checks that the moisture is distributed through the volume rather than concentrated near the injection point or condensing on the coldest surface.

Air movement: internal circulation

Circulation is the mechanism that carries ethylene oxide and moisture into the load and carries residuals out again. The test confirms the air-change rate and the duct velocity against the design intent. A chamber can pass every pressure and temperature test and still sterilise unevenly if the fan runs in the wrong direction or a damper is mis-set, which is why the test is performed with the chamber empty and the duct accessible.

The full cycle

The full-cycle evaluation is the integration test. The controller runs a complete recipe — preconditioning, vacuum, humidification, gas admission, exposure, removal, washing and aeration — without manual intervention, while the witness watches valve states, sensor traces and alarm logs. What matters is not that the cycle completes, but that it completes in the intended sequence, at the intended set points, with no unexplained jumps in any channel.

Reading a Leak Rate Correctly

Pressure decay is a temperature measurement in disguise. In a sealed chamber, a temperature change of 1 °C moves the internal pressure by roughly 0.34 kPa. A test that starts before the chamber has stabilised, or that runs in a room whose temperature swings overnight, will therefore report a leak rate that is partly thermal drift and partly real leakage. The protocol has to require a stabilised starting temperature above ambient, and the calculation has to subtract the temperature effect before the rate is compared with the acceptance limit.

This is the single most common technical argument in FAT review, and it is the same correction that governs the chamber leak test performed on site. A supplier reports a passing rate; the buyer’s engineer notices that the chamber cooled by two degrees during the hold, which accounts for most of the pressure change. The test is only meaningful when the correction is shown alongside the raw start and end values. Ask for the calculation, not just the conclusion.

What the FAT Report Must Contain

The report is the deliverable that outlives the machine. Two years later, when a notified body asks how the chamber was qualified, nobody will remember the test; they will read the file. The documentation requirements are therefore part of the test scope, not an administrative afterthought.

Each attachment should be issued as a PDF with a title that describes its function, the manufacturer’s name and logo, and the equipment number. File names should match the test, so that a future auditor can find the positive leak test without opening twelve similarly named files. Trend curves belong in the attachment, not only in the controller history, because controller data can be overwritten or lost when a hard drive is replaced.

The signature block deserves particular attention. A report signed only by the tester is a record of activity; a report signed by a reviewer and an approver as well is a controlled document. The distinction is exactly what a quality auditor is testing when they ask who accepted the result.

Empty-Chamber Data Is a Baseline, Not a Validation

Every physical test in a FAT is run with an empty chamber. That is deliberate: it isolates the machine from the load and produces a baseline that can be compared between units and over time. It does not prove that a loaded pallet will reach the same temperature, humidity or gas concentration. Heat transfer through a wrapped pallet, the position of the load relative to the circulation path, and the mass of product all change the result, which is why performance qualification is run on the worst-case load rather than on the empty vessel.

Buyers sometimes treat a good uniformity result as evidence that PQ will pass. It is not, and a supplier who presents it that way is either overselling or misunderstanding the standard. The correct reading is narrower and still valuable: the chamber has no gross thermal defect, its sensor map is sound, and the conditions exist for a successful PQ.

From FAT to IQ, OQ and PQ

FAT sits at the start of a chain. Installation qualification confirms that the machine was installed as designed, with the utilities, drawings and documentation actually present on site. Operational qualification repeats the key physical tests in the installed environment, where steam quality, ambient temperature and power supply can differ from the factory. Performance qualification then runs the loaded chamber and produces the evidence that the process, not just the equipment, is capable.

A well-run FAT makes all three cheaper, because the site team inherits a documented baseline instead of starting from zero. A poorly run FAT pushes discovery downstream, where every finding is more expensive and more visible.

Re-validation is triggered by change rather than by calendar alone: a new product family or load configuration, a change to the cycle or the equipment, a major repair affecting the pressure boundary or the control system, and the interval defined in the site’s validation master plan, commonly twelve months. Each trigger is easier to justify when the original baseline is complete.

A Buyer’s FAT Checklist

• Protocol signed before travel: methods, limits, instrument list and sensor map agreed in writing.
• Unit identified by works number: on the protocol, every attachment and the final report.
• Calibration certificates in date: for every pressure, temperature and humidity channel used.
• Seal tested in both directions: positive pressure and vacuum, with temperature correction shown.
• Uniformity map attached: sensor positions numbered and traceable to the report channels.
• Full cycle witnessed, not summarised: valve states and alarm log observed live by your engineer.
• Report set complete: raw data, curves, limits, deviations and three signatures.
• Deviations closed before shipment: each with a root cause, a corrective action and a retest result.

Why Teams Witness FAT at BOCON

BOCON builds ETO sterilizers from laboratory units to two-row industrial chambers, and the FAT is run in-house on the same site where the chamber is assembled. The protocol covers trial operation, the ten physical tests and a full-cycle evaluation, and the report set is issued as signed PDF attachments with trend curves rather than as a summary sheet. Because chamber fabrication, control software and validation support sit inside one organisation, a deviation found during the test is corrected by the team that designed the subsystem, not referred to a third party.

For buyers comparing suppliers, the practical question is not whether a FAT is offered — almost everyone offers one — but whether the protocol is fixed before the visit, whether the data is traceable, and whether the person answering the deviation is the person who can change the design. Details of the industrial EO sterilizer range and the associated station scope are published separately.

Frequently Asked Questions

Reference: ISO 11135:2014 · EN 1422 · ISO 13485. Acceptance limits are defined in the approved FAT protocol for each order.