BESS FAT vs SAT: Test Scope, Records and Acceptance

A battery energy storage system can pass its factory tests and still arrive on site with important acceptance work unfinished. Shipping, field wiring, protection settings, communications and the actual point of connection introduce conditions that the factory could not fully reproduce.

A useful BESS FAT and SAT plan makes that boundary explicit. It identifies what was demonstrated before shipment, what must be verified after installation and what evidence authorizes the next stage. A signed factory report should make site work more focused, while leaving the site's unfinished interfaces visible.

What FAT and SAT should establish

Factory acceptance testing, or FAT, assesses the agreed equipment and functions in the factory configuration before release for shipment. Site acceptance testing, or SAT, assesses the installed configuration and the site interfaces covered by the project. The contract and approved test plan must define the actual scope of both.

Terms vary between suppliers. One vendor may use SAT for basic installed checks while another includes substantial performance testing. Read the test list rather than relying on the acronym. Record where a function was simulated, tested on a representative subsystem or demonstrated with the delivered equipment.

Sandia's energy storage procurement guidance treats factory testing and commissioning acceptance/performance testing as distinct procurement topics. Its commissioning guidance emphasizes adapting the process to the design intent and purchase specifications. The working matrix below follows that project-specific approach; it is not a universal acceptance standard.

Make a requirement-to-evidence matrix

Every acceptance item needs a source requirement, a test method, a boundary and an acceptance criterion. Add who performs the test, who witnesses it when required, which records are retained and who authorizes progression.

A row marked “Passed at FAT” should link to the actual result and configuration. If the installed firmware, protection setting or relevant hardware changes, the responsible team must decide which evidence remains applicable and what requires retest.

Define responsibility before booking the test date

A typical coordination plan involves the equipment supplier, system integrator, installation contractor, commissioning lead and owner's representative. Grid and emergency-response interfaces can add other parties. Their responsibilities must be assigned by the project; the names below are planning prompts, not a contractual allocation.

For each item, identify the person who supplies the procedure, provides the instrument, controls the equipment, records the result, resolves a failure and approves the next stage. “Vendor to attend” is not enough if no one owns the site communications test or the protection-settings record.

Agree on witness notice periods and what happens if a witness cannot attend. Also agree on the evidence needed for a deferred test. A missing witness and a failed function are different issues and should not be recorded with the same status.

Use hold points that correspond to real decisions

A practical plan can place hold points before shipment, before energization, before first controlled operation and before final handover. Each should list its required evidence and the person authorized to release it.

Before shipment, reconcile the delivered configuration and open factory exceptions. Before energization, establish completion of the required installation checks, protective arrangements, operating procedures and site prerequisites. Before controlled charge or discharge, confirm the necessary permissions, controls, operating limits and emergency arrangements.

The precise checks depend on the system and approved plan. Do not use a generic internet checklist to choose battery test voltages, bypass interlocks or authorize energization. Those actions need the equipment documentation and responsible site personnel.

The BESS commissioning equipment checklist complements this article by mapping measurement tasks to instruments. Here the focus is on which evidence closes each acceptance decision.

Write functional tests as observable sequences

“Check emergency stop” leaves too much undefined. A useful test record states the initial condition, approved stimulus, expected response, observed response, timestamps, affected equipment and restoration steps. It also states which part of the chain was simulated.

For example, a simulated input may verify controller logic without proving the installed field sensor and wiring. Conversely, verifying a sensor indication does not prove every downstream action. Keep these boundaries visible so separate pieces of evidence can be assembled into an end-to-end result.

Use approved test facilities and OEM procedures for alarms, protective functions and communications failures. Do not create an uncontrolled electrical or battery fault to demonstrate a protective response. Preserve event logs from the relevant controllers and align their clocks before comparing response times.

Document restoration. Temporary test settings, simulations and inhibited functions need a recorded return to the approved operating state before the next release.

Define the boundary for power and energy tests

A capacity statement without a measurement boundary is difficult to evaluate. Specify whether energy is measured at the battery DC terminals, PCS AC terminals or the site's point of connection. State whether auxiliary loads are included and how they are treated.

Also define initial and final state conditions, operating power, temperature conditions, conditioning requirements, meter arrangement, sample interval and exclusions. The approved contract or method supplies the acceptance values; the test report supplies the actual conditions.

Consider a deliberately simple arithmetic example: a test records 1,000 kWh entering and 900 kWh leaving the same defined AC boundary. The ratio is 90% for that test boundary and period. It should not be compared directly with a DC-only figure or with another test that excludes auxiliary energy. These numbers are illustrative, not an EK performance claim or a BESS acceptance limit.

If a test is interrupted by a site restriction or controller limit, preserve the record and classify the outcome accurately. “Not completed” is more useful than silently shortening the test to produce a passing result.

Match portable instruments to a defined task

A power quality analyzer such as EK-F500 can provide voltage, current, power, waveform and harmonic evidence for an appropriate AC measurement point. Confirm sensors, ratings, uncertainty, synchronization and export requirements for the project. A portable survey instrument is not automatically the agreed reference for contractual energy acceptance.

The EK-C366A DC Resistance Tester may be considered for approved resistance checks on suitable isolated connections. Its test current and connection method must match the asset procedure. Never connect a resistance source across an energized battery circuit.

The EK-H755 Battery Internal Resistance Tester uses a listed 1 kHz AC method and has a published maximum DC input of 100 V between its measurement terminals. It is not a direct high-voltage BESS-string tester. Terminal-to-earth conditions, common-mode exposure, test leads, battery configuration and OEM permission must also be checked; a low voltage difference alone does not establish safe application.

Battery internal resistance is also not a substitute for an agreed capacity or energy test. It can form part of a condition record under a suitable procedure, but should not be converted into a universal remaining-capacity percentage.

Handle exceptions without losing the release boundary

Use an exception log with the requirement, failed or deferred test, affected function, evidence, responsible party, corrective action, target date and retest status. Give each item an identifier referenced by the associated test sheet.

Classify whether the open item blocks shipment, energization, operation or final handover under the project's agreed process. The authorized project parties must approve any conditional release and its restrictions. An open document-format issue and an unresolved protective-function issue should not be treated as equivalent simply because both appear on a punch list.

After correction, record what changed and repeat the affected tests. Consider related functions where the change can have consequences. A screenshot of a cleared alarm is not necessarily enough to close the original failed sequence.

Build the handover package around evidence

The final package should let a future operator identify the installed system and understand the basis of acceptance. Include the approved test plan, executed results, configuration and firmware records, drawings, instrument identities, raw measurement files, resolved exceptions, authorized outstanding items and operating documentation.

Keep raw records alongside reports. A PDF plot is useful, but later questions may require the original event data, units, time basis and configuration. Use consistent equipment names across factory records, site sheets and the asset register.

Training and emergency-response arrangements also need a completion record appropriate to the project. A test package can be technically complete while the receiving operations team is still missing the information required to operate the plant.

Questions to settle before accepting a FAT report

• Was the delivered equipment tested, or was some evidence obtained from a representative configuration?
• Which site interfaces were simulated?
• What changed after testing, and what was retested?
• Which functions and performance checks remain for SAT?
• What evidence is required before shipment, energization and operation?
• Who can accept exceptions and authorize each release?

A well-defined FAT/SAT boundary reduces repeated work and exposes missing work early. When specifying portable test instruments, provide the exact measurement tasks and electrical conditions. When accepting the BESS itself, use the approved project criteria and the evidence from the installed system.