Chiller Evaporator Approach: Diagnose Fouling Without Misreading the Trend

6 days ago

4 min read

Illustration of a water-cooled chiller evaporator with insulated pipe connections.

Start with a repeatable thermal comparison

A rising evaporator approach deserves investigation when the chiller is operating at comparable load, water flow and leaving-water temperature. A single high reading cannot identify the deposit, locate the resistance or establish the maintenance action. The useful question is whether the evaporator now needs a larger temperature difference to transfer the same cooling duty.

This workflow concerns the water-to-refrigerant heat exchanger inside a chiller. Record the chiller model, refrigerant, evaporator construction, active circuits and clean commissioning baseline before comparing readings. Flooded, falling-film and direct-expansion machines can have different normal approaches and diagnostic conventions.

Define approach at the evaporator boundary

For the convention used here, approach equals leaving chilled-water temperature minus refrigerant evaporating saturation temperature: A = Tleaving − Tevap,sat. The water-side temperature drop is a separate quantity, Tentering − Tleaving. With water leaving at 7.0°C and evaporating refrigerant at 4.0°C, approach is 3.0 K. Temperature differences have the same numerical value in kelvin and degrees Celsius.

Trane’s RTAF operating manual exposes evaporator approach, water temperatures and refrigerant saturation temperature as separate operating values. Use the manufacturer’s refrigerant-temperature reference, pressure location and conversion convention. A suction-line surface temperature includes effects that can make it unsuitable as the saturation-temperature input.

Validate measurements before interpreting a change

Check water probes against a suitable reference and verify immersion, thermal contact, insulation and point mapping. Ask a qualified refrigeration technician to validate the pressure transducer and pressure-to-temperature conversion. For refrigerant blends, follow the manufacturer’s dew/bubble or other stated reference; do not silently substitute one saturation convention for another.

Log simultaneous entering and leaving water temperatures, evaporator flow, refrigerant saturation temperature, compressor loading and condenser-side conditions. Select stable windows after startup and staging transients have cleared. Use comparable operating bins rather than comparing yesterday’s low-load snapshot with today’s design-load operation. Preserve the raw data alongside any rolling averages.

As an uncertainty screen, assumed independent standard uncertainties of 0.10 K for leaving water and 0.15 K for refrigerant temperature combine to √(0.10² + 0.15²) = 0.18 K for approach. This is an illustrative propagation calculation, not a universal alarm threshold; correlated errors, drift and calibration bias need separate treatment.

Worked example: the same duty requires a larger driving force

Assume clean water, steady operation, 20.0 L/s, density 1,000 kg/m³ and specific heat 4.18 kJ/(kg·K). Both comparison windows have entering water at 12.0°C and leaving water at 7.0°C. Cooling duty is therefore 20.0 × 4.18 × 5.0 = 418 kW. These rounded fluid properties are example assumptions.

In the clean baseline, refrigerant saturation temperature is 4.0°C. The terminal temperature differences are 8.0 K and 3.0 K. The log-mean difference is LMTD = (8 − 3)/ln(8/3) = 5.098 K. Apparent conductance UA = 418/5.098 = 82.0 kW/K.

In the later window, refrigerant saturation temperature is 2.0°C. Approach rises from 3.0 K to 5.0 K. The terminal differences become 10.0 K and 5.0 K, giving LMTD = 5/ln(2) = 7.213 K and apparent UA = 57.9 kW/K. The inferred conductance has fallen about 29.3%.

This original screening calculation uses Q = UA × LMTD, the heat-transfer framework described in ASHRAE’s heat-exchanger chapter. It assumes approximately uniform refrigerant saturation temperature, negligible heat loss and unchanged geometry. Significant glide, pressure drop, superheat regions or refrigerant maldistribution require a more detailed model.

Matched 418 kW evaporator snapshots with water at 12°C entering, 7°C leaving and 20 L/s. Refrigerant saturation falls from 4°C to 2°C, approach increases from 3 K to 5 K, and apparent UA decreases from 82.0 to 57.9 kW/K.

Original illustrative calculation. The 29.3% apparent-UA reduction supports investigation at matched operating conditions; it does not identify a fouling mechanism.

Separate a thermal symptom from a confirmed cause

The calculation supports a loss of effective heat-transfer performance under the stated assumptions. It does not produce a material fouling factor or prove waterside scaling. Johnson Controls’ plant fault guidance identifies fouling or blocked evaporator tubes as possible causes of excessive evaporator approach.

Investigate water-treatment history, debris, strainers and evaporator pressure drop at matched flow. A restriction may increase hydraulic resistance; a thin insulating film may create a thermal penalty without an obvious pressure-drop change. Refrigerant distribution, charge or level problems and oil-related effects also need assessment by the equipment specialist. A higher approach alone cannot distinguish these mechanisms.

Flow must remain comparable. Trane’s variable-primary-flow discussion explains that reducing evaporator flow changes the water-side heat-transfer coefficient. An apparent-UA comparison across materially different flows can therefore attribute a normal operating change to deterioration.

Close the investigation with a verified baseline

Use the evidence to agree an inspection or cleaning plan with the manufacturer and water-treatment specialist. Follow approved isolation, pressure-release, tube-cleaning and chemical-compatibility procedures. Avoid prescribing an acid concentration or brush type without knowing tube material, enhancement and deposit chemistry.

After corrective work, repeat the same stable operating window. Record approach, duty, apparent UA, water pressure drop and electrical input on the same measurement boundary. Preserve the accepted results as a dated baseline. Set future alerts from the machine’s documented operating map and measurement uncertainty, with persistence rules that exclude normal transitions.