Feedwater heaters quietly account for roughly a third of a modern plant's thermodynamic cycle efficiency, which means a tube leak is never just a small maintenance item. Left undetected, a leaking tube overloads downstream heaters beyond their rated steam flow and can cascade straight into a forced outage. The good news is that a leaking tube almost always announces itself well before failure, through a predictable pattern of shifting level control valve position, rising terminal temperature difference, and creeping feedwater flow, if someone is actually watching those signals together instead of one at a time. Book a demo to see how iFactory correlates those signals automatically the moment a heater starts drifting off its normal operating envelope.
Feedwater System · Tube Leak Detection · Heater Reliability
Feedwater Heater Tube Leak Detection: Catching the Signal Before the Forced Outage
Level control drift, TTD creep, and rising boiler feed pump amperage are not separate anomalies. Together they are the earliest reliable signature of a feedwater heater tube leak, and iFactory watches all three at once.
~33%
Share of thermodynamic cycle efficiency feedwater heaters typically contribute
$500K-$2M
Typical cost range of a forced outage caused by an undetected tube leak
4 signals
Core performance indicators that shift together when a tube starts leaking
2 tubes
Found on inspection after one utility caught rising valve position and pump amps
The Detection Signature
Four Signals That Move Together When a Tube Starts Leaking
No single reading proves a tube leak on its own. What actually confirms one is the combination — several performance indicators drifting in the same direction over the same window of time, which is exactly the pattern documented across real coal-fired and combined cycle tube leak events.
Level Control Valve Position
To maintain a stable heater level as leaking feedwater adds volume on the shell side, the drain level control valve opens further than its normal position for that load point, often the very first indicator to shift. Comparing today's valve position against yesterday's at the same load point, rather than watching an absolute alarm threshold, is what actually surfaces this drift while it is still small enough to investigate calmly.
Terminal Temperature Difference
TTD compares the saturation temperature of extraction steam to the feedwater outlet temperature. A rising TTD against design baseline points to either tube fouling or the reduced heat transfer surface caused by tube plugging. Because both fouling and plugging move TTD in the same direction, this signal is more useful as a corroborating data point than a standalone leak alarm, which is exactly why it needs to be read alongside the other three signals rather than on its own.
Feedwater Flow and Pump Amperage
Internal leakage from the tube side to the shell side effectively bypasses part of the feedwater flow, so flow rate and boiler feed pump motor amperage climb even though unit load has not changed. This is frequently the signal that finally gets attention on a control room screen, since pump amperage approaching its rated maximum is hard to dismiss as noise, but by the time it moves this much the leak has usually already grown well past its earliest stage.
Drain Cooler Approach Temperature
Abnormal drain cooler approach values point to level control problems consistent with a leak adding unexpected volume to the drain cooling section, particularly when paired with a shifting DCA trend. DCA is often the most overlooked of the four signals precisely because it moves more subtly than valve position or pump amperage, which is where automated correlation adds the most value over manual shift-log review.
From First Drift to Confirmed Leak
How a Tube Leak Actually Progresses Through the Feedwater System
A feedwater heater tube leak rarely goes from healthy to catastrophic overnight. Real events documented across coal and combined cycle plants follow a recognizable progression, which is exactly what makes early detection possible if the right data is being watched.
1
Pinhole or hairline crack initiates at a tube wall weak point, typically from corrosion, erosion, or vibration fatigue at a support location, with essentially no measurable performance impact yet.
2
Level control valve position begins drifting as the control system compensates for the small added volume, often the earliest signal available and the easiest to miss on a manual shift log.
3
Feedwater flow and pump amperage climb toward maximum as internal bypass leakage grows, a pattern that has directly preceded confirmed multi-tube leaks in documented plant events.
4
Heater isolation and shell-side drain test confirms tube-side leakage once shell drains are opened after steam and drain connections are sealed, and the affected tubes get located and plugged.
iFactory Watches All Four Signals Together, Automatically.
Level control drift, TTD creep, and pump amperage climb rarely get connected on a manual shift log until the leak is already severe. iFactory correlates them the moment the pattern starts.
Why Tubes Fail in the First Place
The Root Causes Behind Most Feedwater Heater Tube Failures
Detection matters most when a plant also understands why tubes fail, since the same monitoring data that catches a leak early can also reveal which degradation mechanism is actually driving it, and that answer changes what the long-term fix should be.
Erosion at Support Locations
High-velocity extraction steam impinging on tubes near baffle supports gradually thins the wall until it fails, a mechanism that tends to repeat at the same support locations across a tube bundle.
Corrosion Fatigue From Cycling
Units that cycle load frequently subject tubes to repeated thermal stress that combines with feedwater chemistry to accelerate crack initiation, similar in mechanism to the cracking seen in deaerator shells.
Impingement From an Adjacent Leak
Once one tube starts leaking, the resulting jet can erode neighboring tubes if the heater is not isolated quickly, turning a single-tube event into a multi-tube failure within a short window.
Confirming the Leak
Isolate, Test, and Locate: The Confirmation Sequence
Performance signals point to a probable leak, but confirming it and finding the exact tube requires isolating the heater and running a physical test sequence before any repair work can start.
Isolate
Remove the heater from service before damage spreads to adjacent tubes
Prevents impingement erosion on neighboring tubes from the leaking jet
Seal and drain
Seal steam and drain connections, then open shell-side drains
Confirms tube-side leakage is present before further work begins
Air test
Feed a probe through suspect tubes to trace the exact leak position
Locates the leak by depth from the tubesheet face for accurate plugging
Plug and document
Plug the leaking tube and record its position on the tubesheet map
Builds the failure history needed to catch a repeat pattern later
Managing Tube Plugging
Why Plugged Tube Count Is a Trend to Watch, Not a Number to Ignore
Every plugged tube reduces the heater's available heat transfer surface, and that loss compounds. A heater that has quietly accumulated plugs over several outages can end up with a rising TTD baseline that gets mistaken for fouling, when the real cause is simply less tube surface doing the work. Tracking plugged tube count against performance trend, rather than treating each plug as an isolated repair, is what lets a plant catch that shift before it forces a heater replacement decision under outage pressure. A heater approaching its plugging limit, typically defined by the manufacturer as a percentage of total tube count, needs a proactive replacement plan well before that limit is reached, since running past it risks both further performance loss and an unplanned failure of a heavily degraded tube bundle.
Documenting each plugged tube's exact position, depth from the tubesheet face, and failure type on a tubesheet map also builds the evidence base for spotting a repeat pattern, the same way tracking crack location by weld zone matters for a deaerator. If plugs are clustering near the same baffle support or inlet nozzle across multiple outages, that pattern usually points to a mechanical root cause such as steam impingement or flow-induced vibration rather than random material variation, and the fix shifts from simply plugging the next tube to addressing the underlying flow condition.
From the Field
What the Data Actually Looked Like Before a Confirmed Leak
One of our high pressure feedwater heater level control valve positions had been creeping open for about a week, and at the same time feedwater flow and boiler feed pump amps were both trending up toward maximum even though load hadn't changed. Individually those looked like normal operational noise. Together they were the exact signature everyone talks about in tube leak case studies. We isolated the heater, ran the shell-side drain test, and found two leaking tubes on inspection. If we'd been watching those trends as a combined pattern instead of four separate charts, we would have caught it several days earlier.
— Operations Engineer, Coal-Fired Generating Station
2 tubesConfirmed leaking on shell-side drain test
~1 weekValve position drift visible before the leak was formally investigated
4 signalsThat would have flagged the pattern days sooner if correlated
Conclusion
The Leak Was Never Silent. It Was Just Spread Across Four Different Charts.
Feedwater heater tube leaks give a plant days of warning in almost every documented case, but that warning only shows up when level control position, TTD, feedwater flow, and drain cooler approach are watched as one connected pattern rather than four separate trends on four separate screens. Plants relying on shift-log spot checks routinely catch the leak only once it has grown large enough that the signal is impossible to miss, by which point isolation and repair are happening under far more schedule pressure than they needed to.
iFactory pulls all four signals into one correlated view and flags the drift pattern automatically, before a small leak forces an unplanned heater isolation. Book a demo to see your own feedwater heater performance trends mapped this way.
Frequently Asked Questions
Feedwater Heater Tube Leaks — Common Questions
What is the earliest reliable sign of a feedwater heater tube leak?
Level control valve position is typically the earliest indicator to shift, since the control system has to compensate for even a small added volume of leaking feedwater on the shell side well before the leak grows large enough to show up clearly in flow rate or pump amperage. Watching valve position trend against the normal position for a given load point, rather than just its instantaneous value, is what actually catches the drift early.
Book a demo to see how correlated trending catches this signal automatically.
How is terminal temperature difference different from a simple tube leak indicator?
A rising terminal temperature difference against the heater's design baseline can indicate either tube fouling or tube plugging reducing available heat transfer surface, so TTD alone does not distinguish a leak from ordinary fouling. It becomes a strong leak indicator only when it moves together with level control valve position and feedwater flow trends rather than in isolation, which is why single-metric monitoring tends to miss leaks that a combined view would catch.
How many tubes can be plugged before a feedwater heater needs replacement?
Manufacturers typically specify a maximum plugging percentage, often in the range of ten to fifteen percent of total tube count, beyond which the heater's remaining heat transfer surface and structural margin are considered compromised. Plants that track plugged tube count as a running trend rather than a series of isolated repairs are far better positioned to plan a heater replacement proactively instead of reacting to a sudden performance cliff.
Why must a heater be isolated immediately once a leak is confirmed?
A leaking tube creates a high-velocity jet that can erode adjacent tubes through impingement, meaning a single confirmed leak left in service risks turning into a multi-tube failure. Isolating the heater from service as soon as practical after a leak is confirmed is standard practice specifically to prevent this collateral damage before the shell-side drain test and tube location work begins.
Can AI-based monitoring reduce false alarms compared to fixed alarm limits?
Yes. Fixed alarm limits on any single signal tend to either trigger too often on normal operational noise or miss slow drift entirely, while correlating multiple signals against each other and against historical load-normalized baselines significantly reduces false positives while still catching the genuine combined-signal pattern that precedes a real leak.
Contact support for help configuring correlated leak-detection alerts for your heater fleet.
Stop Reading Four Feedwater Heater Trends as Four Separate Stories
Correlate level control position, TTD, flow, and pump amperage automatically, and get flagged the moment the combined pattern points to a tube leak.
