FAT vs SAT: What They Are and How to Write the Protocol

A 200,000 dollar conveyorized furnace arrived at a customer site and failed to reach setpoint. The FAT had been waived because the vendor said it was the same model as the previous installation. It was not. The heating element configuration had been changed in the most recent production run. A proper FAT would have caught it at the factory, before freight, before installation, and before the production schedule was disrupted. The equipment went back. Three weeks of lost production time followed.

FAT and SAT protocols exist because equipment failures at installation are expensive and preventable. They are two different tests that answer two different questions, and conflating them creates gaps in the qualification record.

What is a Factory Acceptance Test (FAT)?

A FAT is a commissioning test performed at the equipment vendor's facility before the equipment is shipped to the customer. It verifies that the equipment was built to the agreed specification and performs to spec under controlled factory conditions, before it leaves the vendor's control.

The attendees at a FAT are typically the customer quality engineer, the customer process engineer responsible for the receiving process, and the vendor application engineer. All three need to be present. The quality engineer signs off on the documentation package. The process engineer validates functional performance against the user requirements. The vendor engineer is there to answer questions and address anything that does not pass.

A FAT tests four categories of performance. Functional performance at specified operating parameters: can the equipment achieve and hold the temperature setpoint, pressure setpoint, or speed setpoint it was purchased to achieve? Safety interlock operation: do the interlocks trigger correctly when a process limit is exceeded? Alarm and fault response: do alarms fire at the correct setpoints, and does the equipment respond correctly when a fault condition occurs? Documentation package completeness: are the installation drawings, P&IDs, calibration certificates, maintenance manuals, and spare parts list present and correct?

A FAT is passed or failed. There is no conditional pass. Punch list items are acceptable only under narrow conditions: the item is minor, it is fully documented, and there is a committed resolution date before the equipment ships. An incomplete documentation package is not a punch list item. It is a failure. The FAT does not close until the documentation is complete.

What is a Site Acceptance Test (SAT)?

A SAT is performed after installation at the customer's facility. The purpose is to verify that the equipment performs in its actual operating environment. The factory environment the FAT was conducted in is controlled. The customer's site is not. Actual utility supply, actual floor loading and leveling conditions, actual adjacent equipment, and actual environmental conditions (temperature, humidity, vibration) can all affect performance in ways the factory test could not reveal.

The SAT tests what the FAT cannot. Utility shortfalls show up at the SAT: the compressed air header feeding the new equipment is at 5.8 bar while the equipment requires 6.5 bar, because the existing distribution system was not sized for the additional load. Interface problems with adjacent equipment show up at the SAT: the conveyor speed handshake between the new furnace and the existing quench tank does not match the expected protocol. Environmental differences show up at the SAT: the production floor runs 12 degrees C warmer than the factory test environment, shifting the cooling curve.

The SAT is also where you verify that the installation team followed the vendor's installation documentation. Equipment that was functionally correct at the FAT can be installed incorrectly, leveled inadequately, or connected to the wrong utility circuit. The SAT catches those errors before production starts.

FAT vs SAT: the key differences

Factor	FAT	SAT
Location	Vendor factory	Customer production site
Timing	Before shipment	After installation
Test conditions	Controlled factory environment	Real operating environment
Utilities	Vendor factory utilities	Customer site utilities (actual)
Primary question	Does the equipment meet the spec?	Does the equipment work at our site?
Regulatory weight	Confirms equipment design	Feeds installation qualification (IQ)

The regulatory weight distinction matters. In regulated environments, the SAT record feeds directly into the installation qualification (IQ) report. An IQ without an SAT record is an incomplete IQ. The FAT record, by contrast, goes into the equipment qualification file and supports the OQ by confirming the equipment was built to the spec being qualified.

When you need both, when you can skip one

Skipping the FAT is defensible in a narrow set of circumstances. The equipment is simple commodity hardware with a well-established track record (a standard lab refrigerator, not a custom thermal processing system). There is no customer-specific configuration. The regulatory risk of the process is low. When all three conditions apply, the FAT adds little value and may not be worth the travel cost.

Skipping the SAT is defensible when the FAT conditions closely replicate the installation site. The vendor tested using the same utilities, the same input materials, and the same operating schedule. The equipment is going into a well-characterized environment with no interface dependencies. The site conditions are documented to match the factory test conditions. This situation is uncommon in practice.

Never skip both in a regulated environment. Under ISO 13485:2016, FDA 21 CFR Part 820, and GAMP 5, equipment qualification is expected as part of validation. Both FAT and SAT are expected components of that qualification. Waiving both requires a documented risk justification. If you are in a regulated environment and an auditor asks why both were waived, "the vendor said it was the same as the last one" is not an acceptable answer.

How to write a FAT protocol that does not get rejected

The test scope section defines the boundary of the FAT. Which equipment is being tested? Which customer-specific configuration? If the equipment has variants, which variants will be tested at the FAT and which will be addressed by documented similarity arguments?

The pre-FAT requirements checklist must be signed off before any testing begins. This checklist confirms that the factory utilities match the test requirements, calibration certificates for all test instruments are current, the documentation package is complete and available for review, and the vendor test equipment is calibrated. Testing before these prerequisites are confirmed produces data that cannot be verified. Auditors will ask for calibration certificates dated before the test. If they are missing, the test results are suspect.

Each test case in the protocol needs a specific acceptance criterion written as a pass/fail measurement. The difference between a useful acceptance criterion and a useless one is objectivity.

A bad acceptance criterion: "Temperature control performance is acceptable." Acceptable to whom? At what measurement point? Over what time period? This cannot be verified objectively and will be rejected in protocol review.

A good acceptance criterion: "Setpoint temperature of 730 degrees C achieved within plus or minus 3 degrees C (range: 727 to 733 degrees C) within 15 minutes of initiating the ramp from ambient. Temperature maintained within plus or minus 3 degrees C for a minimum of 60 minutes, confirmed by three independent thermocouple readings at 15-minute intervals, recorded in test data sheet FAT-TC-001." That criterion can be passed or failed without interpretation.

The deviation log template in the protocol defines what gets recorded when a test case fails or produces an out-of-spec result. The fields are: test ID, deviation description, impact assessment (does this affect the equipment's ability to meet the URS?), proposed resolution, responsible party, and due date. Every deviation is documented and closed before the final sign-off is issued.

The final sign-off matrix requires signatures from the customer quality engineer, the customer process engineer, and the vendor application engineer. All three sign the same document. A FAT signed off by only the vendor is not a customer-accepted FAT.

For a complete example of what a controlled acceptance document looks like, see the batch annealing SOP example on this site, which shows the same level of specificity applied to a process procedure. The Aptibot documentation service writes FAT and SAT protocols to this standard for manufacturers preparing for equipment qualification.

Common FAT and SAT failures and how to prevent them

Utility requirements not confirmed before the FAT. The vendor factory runs single-phase 220V. The customer's equipment requires three-phase 480V. The FAT cannot be run. The team flew to the factory, the vendor is not ready, and the schedule slips by three weeks. Prevent this by requiring the vendor to confirm utility availability against the FAT requirements checklist at least two weeks before the test date.

Punch list items still open at SAT sign-off. The SAT was signed off as "accepted with deviations," and the deviation list listed four open items. Six months later, none of them had been closed. The process is running on equipment with four documented deviations from the acceptance criteria, and the SAT record shows conditional acceptance. In an audit, this looks like running a non-conforming process. Prevent this by requiring every punch list item to have a resolution date, and by not issuing the final SAT acceptance until every item is closed.

Calibration records for measurement equipment not included in the protocol package. The test data shows a temperature measurement of 731 degrees C. The auditor asks which thermocouple made that measurement and when it was last calibrated. If the calibration certificate is not in the protocol package, the measurement cannot be verified. Every instrument used during testing must have its calibration certificate recorded in the protocol and attached to the completed report. This applies to vendor instruments used during the FAT and to customer instruments used during the SAT.

A well-written FAT or SAT protocol prevents these failures by requiring confirmation before testing starts. The pre-requisite checklist is not administrative overhead. It is the mechanism that ensures the test you run is the test you can defend. For more on how SOPs and protocols fit together in a controlled document system, see the guide on how to write a manufacturing SOP.