Hydraulic Test Bench Acceptance Test Checklist: FAT, SAT, and Evidence

A hydraulic test bench acceptance test checklist should connect each agreed requirement to a test method, recorded evidence, and an authorized decision. It should cover the delivered configuration, simultaneous pressure and flow capability, measurement performance, controls, data, safeguards, documentation, and site interfaces. A successful demonstration or a folder of calibration certificates does not prove all of those requirements.

Define the factory acceptance test, or FAT, and site acceptance test, or SAT, before testing begins. Record which checks belong to each stage, who supplies the test articles and utilities, and how deviations will be resolved. The checklist below is a project template. It does not prescribe universal test pressures, hold times, tolerances, or safety procedures; those must come from the approved specification and applicable test methods.

Índice

1. Freeze the acceptance basis before the witness test
2. Separate FAT from SAT responsibilities
3. Build an operating-point coverage matrix
4. Verify the complete measurement chain
5. Agree how borderline results will be decided
6. Witness functions, safeguards, and controlled recovery
7. Test software, reports, and data ownership
8. Use a checklist with evidence and disposition
9. Close deviations and complete the handover
10. Preguntas frecuentes

Freeze the acceptance basis before the witness test

Start with a controlled list of documents: purchase specification, accepted supplier proposal, agreed deviations, hydraulic and electrical drawings, measurement-channel schedule, software scope, and test procedures. Record revision identifiers. A later email changing a sensor range or fixture is part of the acceptance basis only after the responsible parties approve and incorporate it.

Identify the bench and its supplied accessories by serial number or another unambiguous identifier. List the devices under test, or DUTs, that the contract covers, together with adapters, fixtures, software recipes, and exclusions. A pump-testing demonstration does not establish cylinder-testing capability merely because the bench has compatible-looking ports.

Write the acceptance condition for each requirement in advance. “Stable pressure,” “accurate flow,” and “easy reporting” need defined measures, conditions, and evidence. Specify where pressure is measured, which flow is reported, and the period over which stability is assessed.

This record develops the requirements established in the hydraulic test bench buyer’s checklist. If the purchase documents leave an acceptance criterion unresolved, close that gap before the witness session. Changing a limit after seeing a result makes the decision difficult to defend and can conceal a genuine capability shortfall.

Separate FAT from SAT responsibilities

FAT normally evaluates the configured bench at the supplier's facility under agreed conditions. SAT addresses installation and operation at the user's site. The actual division is contractual and technical: neither label automatically defines every test that must occur.

Stage	Typical purpose	Boundary to record
Pre-FAT review	Confirm documents, methods, fixtures, and readiness	Which unresolved items prevent testing?
FAT	Demonstrate agreed functions and capability before shipment	Which site conditions are represented or simulated?
Delivery inspection	Check identification, damage, completeness, and preservation	What changed or became disconnected during transport?
SAT	Verify installation, utilities, interfaces, and agreed repeat checks	Which FAT results remain valid after installation?
Handover	Transfer accepted configuration and operating information	What restrictions or open items remain?

For example, factory cooling water may differ from the site's available supply. A successful factory thermal test then needs an explicit explanation of its applicability. Site electrical supply, network access, foundations, fluid filling, and external guards can also affect the delivered system.

Assign an owner to each interface. Record who provides reference instruments, approved DUTs, consumables, witness personnel, and disposal arrangements. The purpose is to make each test executable without improvising equipment or conditions on the day.

Build an operating-point coverage matrix

Acceptance should demonstrate the required operating envelope, not just separate maximum-pressure and maximum-flow displays. Record pressure, flow, temperature, duration, relevant speed or load, and which channels must operate together. Include the operating modes that matter to the buyer's actual test work.

Test case	Condition to specify	Evidence to retain
Required loaded operating point	Simultaneous pressure and flow at named ports	Time-stamped pressure, flow, and temperature records
Low-flow or leakage measurement	Measurement range and stabilized fluid condition	Results and uncertainty appropriate to that range
Repeated production cycle	Sequence, recovery time, and agreed test period	Cycle records, temperatures, alarms, and interruptions
Alternative DUT or fixture	Mechanical and hydraulic configuration	Fixture identity, recipe revision, and method
Site-dependent operation	Available utilities and environmental conditions	Recorded site inputs and observed capability

A low-flow leakage requirement may need a different measurement range from the main supply measurement. Similarly, an endurance task may impose thermal demands that a short functional demonstration never reaches. Refer to pressure, flow, and power requirements when defining this matrix.

Do not invent a worst-case point by combining incompatible maximum ratings. Select cases from the approved performance envelope and intended use. Where a required case cannot be demonstrated, record the limitation and agreed alternative evidence rather than silently substituting an easier test.

Verify the complete measurement chain

Check each installed channel from sensing element through signal conditioning, data acquisition, scaling, display, and exported result. Match serial numbers and ranges to calibration records. Verify units, sensor type, channel identification, and any correction factors used by the software.

NIST explains metrological traceability as a property of a measurement result supported by a documented calibration chain, with uncertainty contributions. A label saying “NIST traceable” does not by itself establish that a particular bench result is suitable for its intended acceptance decision. NIST metrological traceability guidance.

Review whether calibration conditions and ranges represent the intended measurement. A main-flow meter's certificate may not support a small leakage measurement below its characterized range. Temperature, viscosity, installation, zero behavior, and acquisition settings can matter even when the sensor certificate is current.

Include repeat measurements where the agreed procedure calls for them, and preserve the individual results. Repetition helps reveal variation but cannot by itself remove a consistent bias. When readings disagree, investigate the setup and method before accepting an average. The related guide to inconsistent hydraulic test results explains why apparently similar tests may not be comparable.

A sensor replacement or software scaling change after FAT can affect the measurement chain. Define the evidence and repeat checks required after such changes, including changes made during transport or site commissioning.

Agree how borderline results will be decided

A numerical tolerance and a displayed value are not always enough to decide conformity. State the agreed decision rule, including how measurement uncertainty is considered and what happens when evidence is insufficient. Avoid leaving that choice to the operator after a marginal result appears.

ILAC's G8 guidance addresses decision rules and statements of conformity. Its existence supports making the decision method explicit; it does not supply a universal rule for every hydraulic test bench purchase. ILAC G8 publication information.

Consider an original, dimensionless illustration. Suppose an agreed upper specification limit is 100.0, the measured value is 99.7, and reported expanded uncertainty is 0.5, with its coverage information documented. A simple comparison of the measured value with the limit would accept it. A separately agreed conservative rule requiring measured value plus that uncertainty to remain below the limit would not: 99.7 + 0.5 = 100.2.

This example demonstrates why the rule matters; it is not a required guard band or a hydraulic setting. The parties must select a suitable rule and understand its consequences. “Not accepted under this rule” also does not automatically prove the underlying true value exceeds the limit.

Keep original results when a retest is needed. Record why the retest occurred and whether the setup, method, or equipment changed. Repeating tests until one result passes is not a reliable acceptance strategy.

Witness functions, safeguards, and controlled recovery

Use the approved functional sequence to verify commands, modes, permissives, alarms, and responses. Identify the expected state before each check and the evidence that will demonstrate the required response. Where practical, relate each function to an agreed requirement identifier.

Pressure testing involves stored energy and potential component failure. HSE GS4 emphasizes a planned approach to pressure-test hazards and protective arrangements. The project's competent personnel must determine the test procedure, containment, access control, and other measures appropriate to the equipment. HSE pressure-testing guidance.

This checklist does not authorize bypassing interlocks, opening pressurized circuits, deliberately failing fittings, or creating uncontrolled faults. Safety-function validation requires its own approved methods and responsible specialists. A demonstration that an emergency-stop button changes a screen indication is not sufficient evidence of the machine's complete response.

Document controlled recovery after an interruption: what status the system reports, which results become invalid, and what authorization is needed before the next cycle. Include the behavior of partially completed tests. A report must not show an interrupted test as a normal completed pass.

Different DUTs bring different fixtures, stored energy, and measurement objectives. Use the correct method for the hydraulic component being tested, rather than carrying one acceptance routine across every bench configuration.

Test software, reports, and data ownership

Record the software version, recipe revision, configuration files, and relevant instrument settings used for acceptance. Make a representative test record and follow it through display, storage, export, and retrieval. Confirm that the exported file contains the information needed to understand the result without relying on the operator's memory.

Useful fields include bench and DUT identification, operator or run identifier, date and time with time-zone convention, units, channel names, test conditions, method revision, acceptance limits, decision rule, result, and exceptions. Include raw or suitably detailed time-series data when required by the specification.

Check whether the report uses the same scaling and units as the live display. Confirm treatment of invalid channels, interrupted tests, missing samples, and excluded data. A polished PDF can conceal an incorrect calculation or an unrecorded change of recipe.

Agree on access rights, backups, export formats, and recovery responsibilities. If remote support is included, define its approved access process. Demonstrate restoration or retrieval to the extent agreed in the contract; do not infer reliable long-term record retention from a single successful screenshot.

The operating overview in how a hydraulic test bench works provides context, but acceptance should verify the exact delivered software and data path.

Use a checklist with evidence and disposition

The following template is an original starting point. Replace each bracketed field with project information. Mark a check complete only when its evidence has been reviewed; “not applicable” needs a recorded reason.

Check	Evidence or record	Disposition
Approved requirement and revision identified	[Document and clause]	[Pass / fail / open / N/A with reason]
Delivered bench and accessories match scope	[Serials, inventory, drawings]	[Result and reviewer]
Fixtures and DUT interfaces verified	[Fixture revision and inspection]	[Result and reviewer]
Required operating points demonstrated	[Run IDs and conditions]	[Result and reviewer]
Measurement channels verified	[Channel list and calibration records]	[Result and reviewer]
Uncertainty and decision rule agreed	[Method and rule identifier]	[Result and reviewer]
Functional and safety evidence reviewed	[Approved procedure and validation record]	[Result and reviewer]
Reports and data export checked	[Example files and configuration]	[Result and reviewer]
Site utilities and installation checked	[SAT records]	[Result and reviewer]
Documentation and training delivered	[Document register and attendance]	[Result and reviewer]
Open issues assigned and retested	[Issue register and closure evidence]	[Result and approver]

Add project-specific checks rather than compressing everything into one general “bench works” entry. A requirement covering multiple modes may need several evidence records. Conversely, avoid recording an inapplicable check as passed simply to fill the sheet.

Retain the checklist alongside the evidence package. File names, run identifiers, and revision references should let another reviewer reproduce the reasoning that led to the acceptance decision.

Close deviations and complete the handover

Give every open issue a unique identifier, description, affected requirement, owner, proposed correction, due date, and closure evidence. Identify which issues block shipment, installation, or use, according to the agreed project process. Minor documentation corrections and unverified protective functions should not be treated as equivalent issues.

If the project permits conditional acceptance, state the exact remaining restrictions and approving authority. A supplier's promise to correct an issue is not evidence that the correction has been implemented. Retest the affected functions and related interfaces as needed, and retain the earlier result for traceability.

The handover package should include the accepted configuration, drawings, operating and maintenance instructions, measurement records, software and recipe versions, spare-parts information, training records, and the signed issue register. Confirm that site personnel can locate these documents.

Close the process with an explicit decision by the authorized parties. The checklist organizes technical evidence; it does not itself grant regulatory approval or replace the project's commercial acceptance terms.

Preguntas frecuentes

Does FAT eliminate the need for SAT?
No automatic rule applies. Site installation can change utilities, interfaces, and measurement connections. Agree which checks remain valid and which must be repeated or completed on site.

Are calibration certificates enough to accept measurement accuracy?
No. They are part of the evidence. The installed measurement chain, operating conditions, uncertainty, and decision method must also support the intended result.

Must every bench use the same proof pressure and hold time?
No. Those values depend on the equipment, procedure, and applicable requirements. A general checklist must not invent them.

What happens if a required test cannot be completed?
Record it as open or otherwise unresolved under the agreed process. Assign responsibility and specify acceptable completion evidence before an authorized acceptance decision.