The Efficiency Dropped Over Time: A Diagnostic Decision Tree

Why this matters

Gradual decline is sneaky. The system worked fine for years, so no single day looks like a failure, but the customer slowly pays more and gets less. Because there is no sudden break, the temptation is to call it "old equipment" and move on. Often it is a recoverable loss: a fouled surface, drifting calibration, or a wearing part you can restore for far less than replacement. Knowing how to separate true wear-out from reversible degradation is what makes you worth calling.

Start here: establish the baseline and the slope

You cannot diagnose a drop without a starting point.

• Get a reference value. Commissioning data, the nameplate rating, last year's readings, or the customer's own utility history. Any anchor beats guessing.
• Was the decline gradual or stepped? A slow slide points to fouling, scale, or slow wear. A sudden step down at a specific date points to an event: a part that partly failed, a setting someone changed, or a load that was added. Ask what changed near that date.
• Measure now. Take the same readings the baseline used so you are comparing like to like.

If the drop is a clean step, treat it as a change (see the change-driven decision tree). If it is a gradual slope, continue.

Branch 1: surface fouling and buildup (simplest, most common)

Anything that transfers heat, moves fluid, or filters air loses efficiency as it accumulates film, scale, lint, or debris.

• If heat-transfer surfaces are coated, the system works harder to move the same energy across a dirtier barrier. Clean and re-measure.
• If filters, strainers, or screens are loaded, flow is restricted and the system fights the restriction every cycle.
• If scale has built up in water-side passages, capacity drops and pumping or heating effort rises. Descale or treat the water.

Reversible degradation lives here. Always rule it out before condemning a component.

Branch 2: a wearing component drifting out of spec

Parts do not always fail outright. They drift.

• Check moving parts for wear. A worn bearing, a slipping drive, a stretched belt, or a motor pulling above rated draw all add friction that grows over time.
• Check anything that meters or seals. A valve not closing fully, a seal weeping, a worn nozzle or orifice. Small losses compound.
• Compare actual draw or flow to rating. A reading creeping above nameplate confirms added load from wear.

If a part reads out of spec, restoring or replacing it recovers the loss.

Branch 3: calibration and control drift

The hardware is fine but the brain has wandered.

• Verify sensors against a known reference. A temperature, pressure, or level sensor reading a few units off will run the system inefficiently while looking normal.
• Check setpoints and schedules. Settings drift when people adjust them or when a control loses configuration. A schedule that no longer matches occupancy wastes runtime.
• Recalibrate or restore defaults and confirm the system responds correctly.

Branch 4: the load grew, not the system declined

Sometimes efficiency did not drop; demand rose.

• Look for added load. New equipment, more occupants, a worse-sealed envelope, harder water, a longer operating season. The system delivers the same but against a bigger task, so per-unit cost looks worse.
• Separate this clearly. This is not a repair; it is a sizing or load conversation.

Branch 5: genuine end of life

If surfaces are clean, parts are in spec, controls are calibrated, and load is unchanged, the equipment may simply be at the end of its service life. Cumulative metal fatigue, irreversible heat-exchanger degradation, or a platform that was never efficient by current standards. Document the measured gap from baseline and lay out repair-versus-replace honestly.

Wrapping up

Restore the cheap, reversible losses first (fouling, calibration, a worn part). Re-measure. If the gap closes, you saved the customer a replacement. If it does not, you have the data to justify the next step.

References

• ENERGY STAR maintenance and equipment-efficiency guidance
• U.S. Department of Energy energy-use and equipment resources
• Manufacturer commissioning data and service intervals
• See related: It Was Fine Until We Changed Something decision tree