EES Acceptance Testing & NFPA 110 Load-Bank/Transfer Tests

Acceptance testing is the moment the Essential Electrical System (EES) stops being a set of drawings and submittals and becomes a verified, code-compliant lifeline. This Article covers the field tests that prove a hospital's emergency power supply system (EPSS) — generators, automatic transfer switches, fuel systems, and the distribution that ties them to life-safety and critical loads — actually performs under load, transfers within required times, and is documented well enough to satisfy the AHJ, CMS, and the accreditor on day one and every year after.

The scope here is the acceptance and periodic performance testing of the EES/EPSS as governed primarily by NFPA 110 (Standard for Emergency and Standby Power Systems), with the branch-structure and transfer requirements of NFPA 99 and NFPA 70 (NEC) Article 700/701/517, and the surveillance regime enforced under the CMS Conditions of Participation and TJC/DNV/HFAP Environment of Care standards. Generator sizing, fuel-supply design, ATS selection, selective coordination, and the EES branch architecture itself are covered by their own Articles; this article assumes those exist and focuses on proving they work.

Why EES Acceptance Testing Is a Distinct, High-Stakes Milestone

In most building types, the standby generator is a convenience load. In a hospital it is a patient-safety system whose failure during an outage can be fatal — ventilators, surgical lighting, infant warmers, telemetry, fire alarm, egress lighting, and critical HVAC all ride on it. Because of that, the EES is held to a far higher proof standard than ordinary commissioning:

It is code-mandated, not optional. NFPA 110 requires an on-site acceptance test before the EPSS is accepted for service, and a documented periodic test program thereafter. NFPA 99 ties the EES branches (Life Safety, Critical, Equipment) to specific transfer-time and connection rules. NEC 517 and 700/701 set the wiring, separation, and overcurrent requirements that the test must validate.
It is permanently surveilled. Unlike a one-time TAB or envelope test, the EES is re-tested monthly and exercised annually for the life of the facility. The acceptance test establishes the baseline against which every future survey is judged.
It gates licensure and accreditation. CMS (via 42 CFR 482 / the Emergency Preparedness and Physical Environment requirements) and the accreditors (Joint Commission EC.02.05.07, DNV, HFAP) cite missing or deficient EPSS testing among the most common — and most serious — survey findings. A weak acceptance package follows the building for years.
It interlocks with life safety and infection control. Generator transfer must coordinate with fire-alarm function, smoke control, elevator recall, and — increasingly — the ventilation that maintains pressure relationships in ORs, AIIRs, and pharmacy. A transfer test is therefore also a partial test of the building's life-safety choreography.

The practical implication for an owner/PMO: the EES acceptance test is a go/no-go event for occupancy, it requires a temporary load source (the load bank), and it must be scheduled, witnessed, and documented as a formal milestone — not absorbed into general electrical close-out.

The NFPA 110 Framework: Class, Type, and Level

NFPA 110 classifies every EPSS along three axes, and the acceptance test must be written against the specific classification the design declares. Getting these right is the foundation of a defensible test.

Axis	What it defines	Typical hospital value
Class	Minimum time the EPSS must run on its own stored fuel at full load, without refueling	Class X where X is hours of on-site fuel — commonly Class 96 (96 hours) for hospitals per state/AHJ overlay; NFPA 110 baseline is shorter, but health-care AHJs and CMS emergency-preparedness expectations frequently drive 96 h
Type	Maximum time allowed from loss of normal power to EPSS power available at the load terminals	Type 10 — power restored within 10 seconds for life-safety and critical loads
Level	Consequence of failure; sets the rigor of the testing/maintenance regime	Level 1 — failure could result in loss of human life or serious injury (always Level 1 for hospital life-safety/critical systems)

A hospital EES is therefore almost universally a Level 1, Type 10 system, with the Class set by fuel storage (commonly 96 hours under health-care AHJ rules). The "10 seconds" of Type 10 is the single most-tested number in the entire acceptance regime: it is the wall the transfer-time test measures against.

NFPA 110 also defines the EPS (the energy source — engine-generator set) versus the EPSS (the whole system including transfer equipment, conductors, and controls). Acceptance testing addresses the system, not just the engine, which is why ATS transfer behavior, load segregation, and alarm/annunciation are all in scope.

What the Acceptance Test Has to Prove

A complete NFPA 110 acceptance test is not a single event but a sequenced set of verifications. The on-site acceptance test (NFPA 110, Chapter 7) must demonstrate that the installed system meets the manufacturer's, the design's, and the standard's requirements. The proof set typically includes:

Cold-start / time-to-accept. With the engine cold (worst case), simulate loss of normal power and verify the EPS starts, reaches operating speed/voltage, and accepts load within the Type-10 window (10 seconds), with the engine cranking and starting on the first attempt.
Single-step full-load pickup. The EPS must pick up the full connected (or rated) load in a single step without exceeding the allowable voltage dip and recovery, and without frequency excursion beyond limits — verified during the load-bank and building-load tests.
Two-hour full-load run. Run the EPS at not less than the larger of the actual connected load or the test load, for a minimum of two continuous hours, recording temperatures, pressures, voltage, frequency, and load at defined intervals (commonly every 15 minutes).
Transfer and retransfer. Each ATS must transfer to emergency on loss of normal, and retransfer to normal on restoration after the programmed time delays, in the correct sequence, with correct time-delay-on-engine-start (TDES), time-delay-on-retransfer (TDNE/TDEN), and time-delay-on-engine-cooldown (TDEC).