Steam Turbine Overhaul: When to Open a Machine, What Gets Inspected, and What Deferring It Really Costs

A steam turbine is the most reliable machine in most plants and the most expensive one to get wrong. It will run for years with nothing more than lubrication and vigilance, which is exactly why the decision to open it is so often deferred past the point where deferral was sensible.

The argument for postponing is always the same and always superficially sound. The machine is running. Output is acceptable. An outage costs production, and an overhaul costs money that could be spent elsewhere this year. Nothing bad has happened.

The argument against is that a steam turbine degrades invisibly. Seal clearances open, and the machine quietly burns more fuel for the same megawatts. Blades erode, and efficiency falls a fraction of a percent at a time. Rotor material accumulates creep damage that no external measurement will ever show you. By the time symptoms are obvious from outside, the cheap version of the repair has usually passed.

This article covers how steam turbine overhaul intervals are actually determined, what each level of inspection covers, what the work typically finds, and how to plan an outage that finishes when it was supposed to.

Why Calendar Time Is the Wrong Interval

Ask how often a turbine needs overhauling and the useless answer is a number of years. The useful answer is a count of equivalent operating hours.

Equivalent operating hours, usually written as EOH, exist because damage does not accumulate evenly with the clock. An hour at steady base load in a stable steam condition is nowhere near as damaging as a cold start, which subjects thick rotor and casing sections to steep thermal gradients while material properties are at their least forgiving. Starts consume life disproportionately. So does running at elevated temperature, running with poor steam chemistry, and operating below minimum load for long periods.

An EOH count therefore weights the calendar. Actual running hours are added to a penalty for each start, with cold starts weighted far more heavily than hot restarts, plus additional penalties for temperature excursions, trips and load transients. Two identical turbines commissioned the same week can accumulate EOH at very different rates: a base-load machine in a sugar mill and a peaking machine that starts most days are simply not on the same maintenance schedule, and treating them as though they are guarantees that one is over-maintained and the other under-maintained.

Original equipment manufacturers publish EOH-based intervals for their own machines, and those intervals are the correct starting point. GE Vernova's inspection and maintenance planning guidance and Siemens Energy's maintenance and overhaul service documentation both structure recommendations around operating hours and start counts rather than years.

The important qualification is that these are starting points, not verdicts. A plant with continuous vibration trending, regular oil analysis, steam chemistry control and periodic borescope inspection has evidence about its machine's actual condition. A plant with none of those has only the calendar, and must therefore follow the conservative interval, because it has nothing else to go on. Condition monitoring does not just catch faults. It buys the right to extend intervals defensibly.

The Three Levels of Inspection

Turbine maintenance is usually organised into three tiers of increasing intrusiveness, and confusing them is a common source of budget disputes.

The minor inspection, sometimes called a running or standard inspection, does not open the casing. It covers the governing and protection systems, valve stroking and freedom of movement, the lubrication system including oil condition and filtration, bearing temperatures and vibration history, the turning gear, instrumentation calibration, and a general external condition survey. It typically needs days rather than weeks and is the outage most plants can absorb without drama. Its purpose is to verify that the systems protecting the turbine still work, which matters more than most operators appreciate: an overspeed trip that has never been tested is an assumption, not a protection.

The intermediate inspection adds partial disassembly. Bearings are opened and inspected, journals measured, alignment checked, couplings examined, and the steam path is surveyed by borescope through available access points. Valves come apart for inspection of seats and stems. This level catches a great deal without the cost and duration of a full opening.

The major inspection, which is what most people mean by an overhaul, opens the casing. The upper half is lifted, the rotor is removed, and the machine is assessed properly: blade condition through every stage, root and shroud integrity, seal and gland clearances measured against design, casing and diaphragm condition, rotor inspection with non-destructive testing on critical regions, bearing journals measured and reconditioned, and full alignment re-established on reassembly. This is a multi-week outage requiring cranes, laydown space, specialist tooling and skilled labour, and it is scheduled months to years in advance. Siemens Energy advises planning approximately eighteen months ahead, which sounds excessive until you have tried to source a replacement diaphragm at short notice.

Running turbines without a documented inspection history? Our maintenance and asset management services cover turbine inspection planning, outage scope definition and condition assessment.

The three tiers of steam turbine inspection and what each covers. Only the major inspection opens the casing. Source: MIMAH engineering analysis of OEM maintenance practice.

What a Major Inspection Actually Finds

Open enough turbines and a familiar catalogue emerges. Understanding it is what allows an outage scope to be defined honestly rather than discovered expensively at week three.

Seal and gland clearance loss is the most economically significant finding and the least dramatic to look at. Labyrinth seals control leakage between stages and at shaft ends. They wear, they rub during transients, and clearances open. Steam that leaks past a seal does no work, so the machine burns fuel to produce steam that bypasses the blades entirely. This shows up as heat rate degradation, and heat rate degradation is expensive at a scale that surprises people: analysis published in POWER magazine puts the annual fuel saving from a 1% heat rate reduction at roughly $700,000 for a 500 MW coal unit at 80% capacity factor. Restoring seal clearances is often the single highest-return item in an overhaul scope.

Solid particle erosion occurs where exfoliated oxide scale from boiler tubing and steam piping is carried into the turbine and blasts the leading edges of high-pressure stages. The damage is worst on control-stage and first-stage blading and is aggravated by cycling operation, which accelerates the shedding of scale.

Water droplet erosion attacks the opposite end of the machine. In the last low-pressure stages, steam is wet, and droplets striking long blade tips at very high relative velocity erode the leading edges over years. Erosion shields exist for this reason and need inspection.

Deposits and fouling build on blading when steam chemistry has not been properly controlled. Silica and sodium compounds carried over from the boiler deposit on blade surfaces, changing their aerodynamic profile and restricting flow area. Stage pressures rise, output falls, and the operator sees a machine that will not reach load. This one is almost entirely preventable through water chemistry, which is why chemistry records are among the first things worth reviewing before an outage.

Creep and thermal fatigue affect the high-temperature sections. Creep is slow, permanent deformation under sustained stress at high temperature; thermal fatigue is cracking driven by repeated heating and cooling cycles. Both are cumulative, neither is reversible, and neither is visible without the machine open and non-destructive testing applied to the regions where they concentrate.

Bearing and journal wear, coupling wear, and alignment drift are the mechanical findings, and they connect directly to what the vibration record has been saying. A machine whose vibration trend has been slowly climbing at twice running speed will usually reveal an alignment or coupling issue when opened. Our guide to vibration analysis for rotating equipment covers how those signatures are read before the casing ever comes off.

Valve problems are the quiet safety item. Stop valves, control valves and non-return valves that have not been exercised can stick. A stop valve that fails to close on a trip is the mechanism behind turbine overspeed events, which are among the few failure modes capable of destroying a machine completely and endangering everyone near it.

The Real Cost of Deferring

The case for deferral is always presented as saving money this year. It is more accurate to describe it as borrowing at an unfavourable rate.

The fuel cost accrues continuously and silently. A turbine with degraded seals and fouled blading may be several percent worse on heat rate than its design point. On a large unit that is a seven-figure annual sum in many fuel markets, spent quietly, month after month, with no line item to draw attention to it.

The capability cost is the megawatts the machine can no longer produce. Restricted flow area and degraded stages mean the unit cannot reach its rated output, which in a constrained plant means either buying power or losing production.

The scope cost is the one that surprises plants most. Damage escalates. A blade with early erosion can often be repaired or replaced individually. Left long enough, erosion products liberate material that damages downstream stages, and a repair becomes a re-blade. A bearing caught on a vibration trend is a bearing. The same bearing left to fail can take the journal, and a shaft repair or rotor replacement moves the cost into a different order of magnitude entirely, with a lead time measured in many months.

The failure cost is the tail risk. Forced outages do not respect the production calendar. They happen at full load, they take out the unit without notice, and they often cause collateral damage. Every plant that has experienced a rotor or blade failure can describe exactly how much it cost, and the number is never close to what the deferred overhaul would have been.

There is a legitimate counter-argument, and it deserves stating fairly: overhauling too early wastes remaining life and introduces risk, because every disassembly and reassembly is an opportunity for error. Machines are damaged by unnecessary maintenance as well as by neglect. The answer is not to overhaul on a fixed calendar regardless of evidence, but to build the evidence base that lets you decide. That is the difference between a condition-based and a time-based strategy, and it is the same argument we set out in preventive versus predictive maintenance.

The four costs of deferring an overhaul, escalating from a continuous fuel penalty to the tail risk of a forced outage. Source: MIMAH engineering analysis.

Planning an Outage That Finishes on Schedule

The engineering of a turbine overhaul is well understood. The schedule performance is not, and the reason outages overrun is almost never technical.

Define the scope before the outage, then freeze it. The single largest driver of overrun is work added after the machine is open. Some discovery work is unavoidable, because you genuinely cannot see inside a closed turbine, but the proportion that is genuine discovery should be small if the pre-outage inspection and condition history were taken seriously.

Plan for discovery deliberately. Because discovery is inevitable, carry a contingency in both schedule and budget, and pre-agree the decision authority. The worst outcome is a crew standing idle while a finding travels up three management levels for approval.

Sort out long-lead items first. Diaphragms, blades, bearings, seals and specialised fasteners have lead times that do not compress for anyone. Order against the anticipated scope, not against confirmed findings, or the outage will finish when the parts arrive rather than when the work is done.

Secure the specialists early. Turbine overhaul labour, balancing specialists, non-destructive testing technicians and machining services are finite regionally, and outage seasons cluster.

Prepare the site. Crane access, laydown area, clean covered space for rotor storage, lighting, scaffolding and lifting equipment certification are all things that are obvious and all things that have delayed real outages.

Record everything. Every clearance measured, every alignment reading, every finding photographed, every NDT result filed. This documentation is what makes the next overhaul planned rather than exploratory, and it is what supports any extension of interval afterwards. It also matters for vibration acceptance on restart, where ISO 20816-2 provides the evaluation framework for large turbines and generators.

Commission properly on the way out. Overspeed protection tested, valves stroked, alignment verified hot as well as cold, vibration baselined at the new condition. A rushed restart is how a well-executed overhaul turns into a trip three days later.

Have an outage coming and no scope definition yet? Our industrial engineering team builds inspection-driven scopes and manages turbine and controls work packages. Our project record includes a major steam turbine generator overhaul for NNPC in Nigeria and a plant audit with four turbine and controls contracts at KRPC.

The recurring findings when a steam turbine casing is opened. Seal and gland clearance loss is usually the most economically significant. Source: MIMAH engineering analysis.

Frequently Asked Questions

How often does a steam turbine need a major overhaul? There is no single figure, which is why manufacturers publish intervals in equivalent operating hours rather than years. The interval depends on the machine's design, its duty, how often it starts and how well its steam chemistry and lubrication have been controlled. A base-load industrial machine and a frequently cycled unit accumulate life consumption at completely different rates. Follow the OEM interval unless you have condition evidence that justifies departing from it.

Can an overhaul interval be extended safely? Yes, but only with evidence. Plants running continuous vibration monitoring, routine oil analysis, disciplined steam chemistry and periodic borescope inspection have grounds to extend, because they can see the machine's condition without opening it. Extending on the basis that nothing has gone wrong yet is not condition-based maintenance; it is hoping.

What does an overhaul actually cost? It varies enormously with machine size, scope and what the inspection finds, so any figure quoted without seeing the machine is marketing. The more useful way to frame it is against the alternative: continued heat rate degradation, lost output capability, escalating repair scope, and the possibility of a forced outage with collateral damage. Those costs are real and they compound.

How long does a major inspection take? Weeks rather than days for the outage itself, with planning starting well over a year ahead for a large machine. Duration depends on machine size, crane and access arrangements, the scope agreed, and how much discovery work emerges once the casing is open.

What is the most valuable single item in a typical overhaul scope? Usually restoring seal and gland clearances. It is unglamorous work with an immediate and measurable effect on heat rate, and heat rate is where the money is on any thermal machine.

Should the OEM always do the work? Not necessarily, and independent engineering firms perform major turbine work routinely. What matters is demonstrable competence on the specific machine class, access to correct drawings and tolerances, proper NDT capability, balancing capability, and documentation discipline. Ask for machines of similar type the contractor has overhauled, and ask to see the documentation package they handed over.

Turbines Reward Evidence and Punish Assumption

The plants that get long, trouble-free service out of their turbines are not the ones that spend the most on maintenance. They are the ones that know the most about their machines: EOH tracked properly rather than estimated, vibration trended rather than spot-checked, steam chemistry controlled rather than sampled occasionally, oil analysed on a schedule, and every previous outage documented well enough that the next scope starts from evidence rather than from a blank sheet.

That knowledge is what makes the overhaul decision straightforward. You open the machine when the evidence says the machine needs opening, you know roughly what you will find before you find it, and the parts are on site because somebody ordered them against a scope that was defined rather than discovered.

The alternative is running to the calendar and hoping, which works right up until the morning it does not.

Have a turbine approaching its interval, or one whose history nobody can produce? Talk to our engineering team. We will review the operating record, the vibration and oil history, and tell you what the machine actually needs before anybody prices an outage.