Repair or Replace? The Question Most Operations Can’t Answer with Data

When the chiller technician says “you’re looking at another repair — probably $8,000 — but this unit is getting toward end of life,” what does the FM do with that?

In most commercial operations, the answer depends on the budget cycle, the ownership group’s appetite for capital expenditure, and whatever institutional knowledge the FM has about that specific unit’s history. Sometimes there’s an informal sense of how many times this unit has been repaired. Sometimes there isn’t.

What’s rarely present: a documented decision framework with defined criteria for when the repair calculus shifts toward replacement — and data on the specific asset that would apply those criteria. The repair vs. replace decision happens dozens of times per year across a commercial portfolio. Without a consistent framework, each decision is made independently, with inconsistent data and inconsistent logic. The aggregate result is a portfolio of assets where some are replaced too early, most are repaired past their optimal replacement window, and a few are run until failure at maximum cost.

This guide gives FM teams five data-based triggers that indicate when replacement should be the serious conversation — and what data to bring to that conversation.

→ How CapEx planning connects to asset lifecycle data: CapEx Timing in FM

Why “It’s Still Running” Is Not a Replacement Criterion

The absence of failure is not evidence of optimal performance. An asset can be running outside its optimal efficiency range, consuming more energy per unit of output than a replacement would, requiring more maintenance labor per year than a new unit would need, and still be classified as “working” because it hasn’t stopped.

The cost of keeping that asset running — not the repair event, but the ongoing cost of operating an aging, inefficient, maintenance-intensive unit — often exceeds the amortized cost of replacement faster than the capital planning conversation reflects.

ASHRAE publishes equipment useful life data for commercial HVAC and mechanical systems that is widely used as a planning baseline. BOMA’s Experience Exchange Report documents maintenance cost patterns by asset type and building age. Both sources point to the same finding: the total cost of ownership for aging equipment increases non-linearly after the asset reaches the end of its design life — with energy cost, maintenance frequency, and failure probability all accelerating simultaneously.

The Five Replacement Triggers

For each trigger, the question is binary: does this condition apply to the asset under review?

If three or more triggers apply: replacement is the recommended capital conversation. If two triggers apply: schedule a formal condition assessment before the next repair. If one trigger applies: document it in the asset record and reassess at next corrective event.

TRIGGER 1 — Age Beyond Useful Life
The test: Is the asset operating beyond the manufacturer’s stated useful life, or beyond the ASHRAE/industry standard useful life for this asset type? Reference benchmarks (ASHRAE):

• Centrifugal chiller: 20–23 years
• HVAC rooftop unit (RTU): 12–15 years
• Air handling unit (AHU): 15–20 years
• Boiler (commercial): 20–30 years
• Emergency generator: 20–30 years
• Walk-in cooler/refrigeration: 10–20 years
• Commercial water heater: 10–15 years
• Fire alarm control panel: 15–20 years

Why it matters: Useful life benchmarks reflect design life under normal operating and maintenance conditions. Assets operating past this threshold have exhausted the reliability design buffer. Parts availability decreases, manufacturer support ends, and failure probability increases.

TRIGGER 2 — Corrective Maintenance Frequency Threshold
The test: Has this asset required 3 or more corrective work orders in any rolling 24-month period?

Why it matters: The 50% Rule (a widely applied FM planning heuristic) states that when annual maintenance cost on an asset exceeds 50% of its replacement cost, replacement should be evaluated. Three corrective events in 24 months, depending on the asset and repair costs, often crosses or approaches this threshold. More importantly: repeat corrective events on the same asset indicate a systemic condition — not random failure — that repair is treating symptomatically without resolving.
How to find this data: Asset-level work order history in your CMMS. If your system doesn’t surface corrective events by asset over time, this data exists in work order records but requires manual extraction.

TRIGGER 3 — Cumulative Repair Spend Threshold
The test: Has the cumulative spend on corrective and non-routine maintenance for this asset in the last 36 months exceeded 40% of its current replacement cost?

Example: An HVAC rooftop unit with a current replacement cost of $22,000 that has received $9,500 in corrective maintenance over the last 36 months has crossed the 40% threshold. The next repair is not just a repair cost — it’s an incremental contribution to a cumulative spend that should have triggered a replacement conversation 12 months ago.
Why it matters: This trigger requires knowing both cumulative spend (asset-level work order history) and current replacement cost (vendor quote or estimating tool). Without both numbers, the trigger can’t be applied.

TRIGGER 4 — Energy Consumption Anomaly
The test: Is the asset consuming materially more energy (>15% above baseline) to maintain the same output as when it was new or last serviced?

Why it matters: Energy efficiency degradation is one of the least visible costs of aging equipment. An HVAC unit that runs 18% longer to maintain the same setpoint than it did three years ago is not a neutral cost — that delta appears in the utility bill every month, every year, until the unit is replaced. For large commercial HVAC systems, this energy penalty can run $5,000–$20,000 per year per unit.
How to detect it: BAS energy consumption data, runtime hours from IoT sensors, or utility bill benchmarking against historical baselines.

TRIGGER 5 — Parts Availability or Manufacturer Support Termination
The test: Have parts for this asset been discontinued by the manufacturer, or has the manufacturer terminated service support for this model/generation?

Why it matters: An asset requiring parts that are no longer manufactured creates two problems: parts cost increases significantly as supply decreases (often 2–4x normal), and lead times extend from days to weeks. An emergency failure on a unit with discontinued parts is an emergency that cannot be resolved quickly — with all the operational consequence that implies.
How to find this data: Direct inquiry to the manufacturer or authorized service contractor. Most vendors know which generations are approaching or past end-of-life support.

Applying the Triggers: A Decision Template

When a corrective maintenance event prompts the repair vs. replace conversation, document the following before making the decision:

Sources:

• ASHRAE — Equipment useful life data by asset type: https://www.ashrae.org
• BOMA International — EER maintenance cost patterns by asset type and building age: https://www.boma.org/BOMA/Resources/EER
• IFMA — Asset management and lifecycle benchmarking: https://www.ifma.org/resources/research-benchmarking/
• McKinsey — Predictive maintenance and CapEx optimization: https://www.mckinsey.com/capabilities/operations/our-insights
• Facilities Dive — Asset lifecycle and commercial building CapEx trends: https://www.facilitiesdive.com