Commissioning persists when seasonal testing, issue closure, operator knowledge, trend review, change control, and recommissioning triggers survive turnover.

Treat occupancy as another operating condition

Pre-occupancy testing may occur under partial loads, temporary schedules, mild weather, incomplete tenant fit-out, limited domestic-hot-water use, or without the operators who will run the building. After occupancy, real use introduces simultaneous heating and cooling, schedule overrides, tenant requests, humidity, ventilation, process loads, domestic demand, after-hours operation, weather extremes, and maintenance decisions that may not have been present during initial tests.

The commissioning plan should identify which systems and modes could not be fully demonstrated before occupancy and define the post-occupancy path. Completion status, conditional acceptance, warranty start, retainage, payment, contractor obligations, owner acceptance, and regulatory sign-off are separate decisions governed by the project documents and responsible parties; one should not be inferred from another.

Finish seasonal and deferred work with the original intent in view

Carry every deferred test, seasonal mode, failed step, pending correction, incomplete trend, training item, and unavailable document into one issue record with a responsible party, due date, prerequisites, required evidence, retest method, and acceptance authority. Examples may include heating at design conditions, cooling and dehumidification, economizer operation, heat recovery, emergency modes, smoke control interfaces, domestic hot-water demand, irrigation, freeze protection, or load-shed sequences.

DOE FEMP guidance recognizes that some commissioning checks occur after acceptance and that seasonal testing may remain required later. The project should preserve access to the design intent, owner requirements, approved sequences, test procedures, prior results, and change history so a later test does not become an isolated vendor demonstration against a different understanding of the system.

Hand over an operating record—not a document pile

The systems record should make it practical to find current sequences, setpoints and reset logic, schedules, alarm and override expectations, equipment and point identifiers, diagrams, balancing information, functional test results, open issues, warranties, maintenance requirements, control backups, trend definitions, training materials, and responsible contacts. Superseded drafts and undocumented field changes should not compete with the accepted record.

Training should be tied to the installed system and normal, seasonal, emergency, startup, shutdown, and troubleshooting conditions. Record who attended, which modes were demonstrated, what could not be shown, which questions remain, and how future staff will be trained. Operator knowledge is part of persistence; a technically correct sequence can still drift when nobody knows why a setpoint, interlock, or alarm exists.

Keep issue closure evidence-specific

“Corrected,” “contractor addressed,” and “monitor” are not complete dispositions. Closure should identify the condition, corrective action, affected equipment or sequence, date, supporting submittal or work record, test or trend evidence, result, remaining limitation, and person authorized to accept it. A changed sequence should be reflected in controls, documents, training, and the systems record—not only in the issue log.

Some results are partial. A failed component may be replaced while a seasonal test remains open; a control edit may pass at one load but require trend review across a full operating week; an owner may accept a known limitation with a management plan. The record should preserve those distinctions so a closed administrative status does not imply a broader technical conclusion.

Use trends and operator feedback as a combined diagnostic

Trend review can test schedules, setpoints, resets, temperatures, pressures, valve and damper response, staging, simultaneous loads, cycling, alarms, energy use, and other performance indicators selected for the system. Point names, units, intervals, time synchronization, sensor quality, overrides, gaps, and data retention should be checked before the graph becomes evidence.

Operator observations, work orders, comfort calls, alarms, maintenance findings, and utility patterns provide context that automation data may not show. DOE’s building re-tuning process combines trend analysis, a building walk-down, corrective action, savings analysis, and continued use in operations. The goal is not an endless dashboard; it is a repeatable route from a meaningful exception to investigation, action, verification, and an updated operating record.

Define what should trigger recommissioning

Triggers may include repeated complaints, unexplained energy or demand change, control-system migration, major tenant or use change, equipment replacement, sequence edits, sensor failure, recurring manual override, large maintenance event, building-envelope work, electrification, capital renovation, persistent alarms, failure to maintain critical conditions, or elapsed time for a high-consequence system.

Assign an owner, review frequency, decision threshold, response path, budget, and evidence-retention period for the persistence plan. Ongoing monitoring, periodic re-tuning, targeted functional testing, and full recommissioning are different levels of effort. The appropriate response should follow the system, consequence, evidence, and operating change rather than a calendar alone.

Primary sources

Sources and technical references.

Source review current as of September 1, 2026.