An HRSG doesn't fail because your team missed something obvious — it fails because a combined-cycle unit runs thousands of thermal cycles a year and the early warning signals (a 3°C metal-temp drift, a 0.2 bar dP shift, a chemistry excursion at 2am) get lost between the DCS, the chemistry log and last outage's UT spreadsheet. Industry studies consistently show that most HRSG tube failures are detectable weeks before rupture — but only if the signals from inspection, chemistry, condition monitoring and prior repairs land in the same place. This guide walks through how to build that layered leak-detection and failure-prevention program using OXMAINT AI, the AI-powered CMMS built for combined-cycle reliability teams.
Catch the HRSG Tube Leak Weeks Before the Forced Outage.
OXMAINT AI, the AI-powered CMMS/maintenance management software, connects the full workflow on one platform — condition alerts and inspection findings in, defects raised and prioritised, corrective work orders assigned, and preventive & predictive PM cadence tuned per HRSG circuit.
The 5 Layers of HRSG Leak Detection — Not One of Them Is Enough Alone
One sensor or one inspection is never enough — leak detection needs five overlapping layers, each catching what the others miss. OXMAINT AI holds every layer's data on the same asset record, so a signal from any one becomes a defect the others can validate. Sign up free and start layering your first detection stream in OXMAINT AI.
Where HRSGs Actually Leak — A Circuit-by-Circuit Map
HRSG leaks cluster in signature zones — HP superheater bends, LP evaporator bends, attemperator sleeves, economiser inlets. OXMAINT AI's circuit-level view separates HP, IP and LP so a signal doesn't get averaged into a meaningless "health score." Book a demo to see the circuit-level view on an HRSG like yours.
The Signal-to-Work-Order Path
Detection without action is just noise — most HRSG programs leak value at the handoff. OXMAINT AI closes the loop: every alert becomes a timestamped defect, defects roll into scheduled work orders, and every WO carries the signal history that raised it. Start free — turn your first HRSG alert into a scheduled WO in OXMAINT AI.
Every Signal That Doesn't Become a Work Order Is a Missed Warning.
OXMAINT AI wires acoustic alerts, chemistry excursions, dP shifts and inspection findings into one defect stream — and one defect stream into one work-order pipeline your outage team can actually plan around.
Cycling Kills HRSGs — Chemistry Discipline Keeps Them Alive
Daily cycling drives 3–5× the thermal fatigue of a baseload unit, and most damage happens in the transient windows — startup, shutdown, layup. OXMAINT AI logs chemistry excursions against unit state so you can see which transients cost you the most tube life. Book a demo to see chemistry-vs-unit-state trending in OXMAINT AI.
| Unit State | Chemistry Risk | Dominant Damage | Where to Watch |
|---|---|---|---|
| Cold start | High O₂ ingress, low pH excursion | Oxygen pitting, FAC initiation | LP evap, economiser inlets |
| Load-following | Cation conductivity swings | Under-deposit corrosion | HP evap tubes, drum internals |
| Shutdown | Attemperator overspray, thermal shock | Thermal fatigue cracking | Attemperator sleeves, SH outlet |
| Wet layup | Air ingress, stagnant water | Oxygen pitting, MIC | Downcomers, feedwater lines |
| Steady load | Steady-state carryover | Steam-side oxidation | SH / RH tube inner surface |
Planned Corrective Maintenance — Turning Detected Leaks Into Scheduled Repairs
The real win isn't detecting the leak — it's holding the unit to the next planned outage instead of forcing a shutdown now. OXMAINT AI bundles the tube history, failure mechanism, repair spec and parts availability into one work-order package ready for outage scheduling. Sign up free and package your next detected defect into a scheduled WO in OXMAINT AI.
What OXMAINT AI Gives an HRSG Reliability Team
OXMAINT AI is built for the combined-cycle reality — three pressure circuits, five detection layers, thousands of thermal cycles a year, one outage window to act on what you've learned. Below are the capabilities that make HRSG leak detection and failure prevention operational. Start free and put your HRSG asset register on OXMAINT AI today.
The acoustic system caught the first hint at 3am — a low-grade signal from HP superheater panel 4 that would have been dismissed as background noise on its own. Because the platform pulled that alert alongside a slow makeup-water rise from the previous week and a UT trend from the last outage that already had that panel on the watch list, the defect opened itself. We rode the unit to the planned outage 11 days later and repaired the tube in the scheduled window instead of on a forced outage.
Frequently Asked Questions
Detect the Leak Once. Repair It on Your Schedule, Not the Unit's.
Move your HRSG leak-detection layers onto OXMAINT AI — acoustic, water balance, chemistry, thermal and inspection signals converging into one defect stream, one work-order pipeline and one repair evidence chain.
-root-cause-analysis-and-prevention-programs.png)
