A commercial office tower ran two 500-ton water-cooled centrifugal chillers. An energy consultant flagged one machine’s specific power well off design and recommended replacing it, and a capital project was already moving. The water treater was asked to explain the chemistry. A structured diagnostic built only on data the chiller already reports found the real problem: biofilm on the condenser tubes, insulating at roughly eight times the fouling the machine was rated to tolerate. The cause was a pump energy retrofit, not the chemistry program. The machine was cleaned, the cause was corrected, and the replacement was cancelled.
A cooling tower can be inspected. A chiller cannot. The condenser and evaporator bundles sit inside a pressure vessel with refrigerant on the other side of the tube wall, so their condition is invisible until someone pulls a waterbox.
An energy consultant flagged specific power on one machine well off design and recommended replacing it. The owner was being asked to choose between two guesses: the mechanical side said “old machine,” and the usual chemistry answer, “the index was high, so it must be scale,” would have been a guess too.
Nobody had trended approach temperature against the machine’s own baseline, so a problem that had been building for fourteen months had no visible history.
Applied the Chillers in the Dark diagnostic method: one stabilized data set from the control panel, compared to the machine’s own commissioning baseline at the same load.
Closed the heat balance first to prove the data could be trusted, then worked the decision trees to separate waterside from refrigerant-side and flow causes before anyone opened the machine.
Quantified the fouling against the AHRI-rated allowance, wrote down the ranked hypothesis, and confirmed it with the cheapest tests that could change the answer, at an outage the site was already having. No lab budget and no teardown to reach the diagnosis.
Physical cause: biofilm fouling with early under-deposit attack. With flow verified and bundle area from the OEM submittal, LMTD rose from 5.26 to 8.55 °F and the overall heat transfer coefficient fell from about 312 to 190 Btu/h·ft²·°F, a 39% loss. That is a fouling factor near 0.0021, roughly eight times the 0.00025 AHRI-rated condenser allowance. At the outage, the tube ends and lower pass carried soft gray-brown slime and no hard crust. Dilute acid produced no fizz, loss on ignition returned a high organic fraction, and shallow pits with black paste under a thin crust showed sulfate-reducing bacteria starting to attack.
Latent root cause: a mechanical decision, not a chemistry failure. Condenser flow had been cut to save pump energy with no review of tube velocity, biocide program, or filtration, and no approach trending that would have caught it in month three instead of month fourteen. The same gate also caught the second machine: its computed flow came back at 969 gpm against a nameplate 1,100. Using nameplate would have inflated its fouling factor by about a third and produced a cleaning recommendation on a machine that only needed its flow checked.
Free PDF Download
Get the full case study
Complete data, methodology, and results — one page.
Independent chiller and cooling-water diagnostics for commercial, institutional, and industrial systems. We read the data your machine already reports and tell you which problem you actually have before you spend money on it. No chemistry sold.