Steam Turbine Repair: Single-Stage, Multi-Stage, and What Each One Involves

Repair and overhaul get used as though they mean the same thing, and the confusion costs real money. An overhaul is a scheduled event driven by running hours: the machine comes apart on a calendar, gets inspected, gets restored to a known condition, and goes back. A repair is driven by a fault. Something has failed, or is measurably on its way to failing, and the work is defined by that fault rather than by an interval.

The distinction matters because it changes almost everything downstream. An overhaul can be planned around a plant outage, budgeted a year ahead, and scoped from experience. A repair is scoped from findings, which means the price quoted before the casing comes off is an estimate and nothing more. Treating one like the other is how repair budgets double.

This article covers what repair actually involves on the machine types most operators run, how the scope changes between them, and what separates a competent repair contractor from one who will hand you a surprise halfway through.

Repair, Overhaul, or Replace

Three questions decide the route, and they are worth answering in order.

Is the machine still fit for its duty? A turbine derated by fouling, erosion or a previous poor repair may be running at a load its owner no longer needs. If the plant's steam conditions have changed since commissioning, the honest answer is sometimes that repair restores a machine nobody wants.

Is the damage confined? Bearing damage, seal wear, governor faults and coupling problems are confined. They live in accessible parts, they have known repair routes, and the machine around them is intact. Rotor cracking, casing distortion and widespread blade-path damage are not confined, and the cost curve rises steeply once the rotor is in question.

Is the part obtainable? This decides more repair jobs than engineering does. A machine whose original manufacturer has withdrawn support, or whose blade sections are no longer produced, is a reverse-engineering job whether or not anybody calls it one.

Only after those three does price mean anything. A quotation for a repair whose scope has not been established is a quotation for the contractor's optimism.

Single-Stage Turbines

Single-stage machines drive pumps, fans, compressors and small generators. They are simple, they are often the most neglected rotating equipment in a plant precisely because they are simple, and they fail in a small number of predictable ways.

The dominant faults are wear rather than catastrophe. Carbon rings and labyrinth seals wear open, and steam that should be doing work leaks past instead. Journal bearings wear or wipe. The governor, frequently a mechanical one on older machines, drifts out of adjustment or sticks. Nozzle blocks erode where wet steam has been passing through them for years. The wheel itself can suffer deposits or erosion, but on a single-stage machine the wheel is rarely the reason the unit came off line.

That makes the repair scope relatively contained. A typical single-stage repair covers rotor inspection and cleaning, replacement of seals and bearings, nozzle block assessment, a governor rebuild or replacement, and a balance check before reassembly. The rotor usually goes to a balancing machine, the casing gets checked for distortion at the joint, and the whole unit goes back with new gaskets and a fresh alignment.

The trap on single-stage machines is treating the repair as a parts swap. Replacing wiped bearings without asking why they wiped returns the machine to service with the cause intact. Bearing damage is a symptom at least as often as it is a fault, and the causes sit in lubrication, alignment, and the vibration the machine was running with before it stopped. Our guide to common steam turbine failures covers the failure modes that masquerade as bearing problems.

How repair scope changes between single-stage and multi-stage machines. The difference is not size, it is how much can be known before the casing comes off. Source: MIMAH engineering practice.

Multi-Stage Turbines

Multi-stage machines are a different proposition. More stages mean more sealing surfaces, more blade rows, a longer and more flexible rotor, and a thermodynamic path where a fault in one stage changes the conditions every stage after it sees.

The failure modes broaden accordingly. Blade-path deposits change stage pressures and load the thrust bearing in ways the design never intended. Solid particle erosion cuts into leading edges in the high pressure stages. Wet steam erodes the last stages of a condensing machine. Diaphragms distort, crack, or foul until stage pressure drops move measurably away from design. Rotors bow, and on machines that have been through a trip and a hot restart, they sometimes bow permanently. Thrust bearings fail because something upstream changed the axial balance and nobody read the pressure survey that would have shown it.

Repair scope on a multi-stage machine is therefore rarely known before disassembly. What can be known beforehand is the evidence: stage pressure readings against design, thrust position trends, bearing metal temperatures, vibration history with phase, and steam chemistry records. A contractor who asks for those before quoting is scoping the job. One who does not is guessing and will reprice later.

The work itself divides into rotor scope and stationary scope. Rotor scope covers cleaning, non-destructive examination of blades and roots, blade repair or replacement, shroud and tenon work, journal and thrust collar restoration, and balancing. Stationary scope covers diaphragm cleaning and repair, seal renewal, casing joint restoration, nozzle work and valve overhaul. Either can dominate the cost, and which one does is not predictable from the outside.

Getting the inspection right is what keeps a multi-stage repair from becoming an open-ended one. The sequence, the hold points, and what each finding actually means are set out in our turbine inspection checklist.

Combined Cycle Machines

Combined cycle steam turbines carry the same mechanical scope as any multi-stage machine plus a duty cycle that changes what fails.

These units cycle. They start, they stop, they follow the gas turbine and the grid, and every start puts a thermal transient through thick casings and rotors that would rather stay at temperature. The result is that thermal fatigue moves up the list of failure modes while steady-state wear moves down. Cracking at casing corners and steam admission areas, thermal distortion at horizontal joints, and rotor surface cracking are cycling damage, not running damage, and a repair scoped without the start count is scoped against the wrong history.

The second difference is commercial. Combined cycle plants sit inside dispatch obligations, and outage windows are short and expensive to extend. Repair scope on these machines is very often a negotiation between what engineering wants and what the outage will allow, which makes the pre-outage inspection and the parts strategy more important than the repair technique.

What a Competent Repair Scope Looks Like

A repair scope worth signing has five parts, and the absence of any one of them is a warning.

A findings-based structure. The scope should separate work that will definitely happen from work that depends on what inspection reveals, with pricing for both. A single fixed price for a repair nobody has inspected is either padded or about to change.

Named hold points. There should be defined moments where work stops, findings are presented, and the owner decides. Rotor examination results and diaphragm condition are the usual two. Without hold points, decisions get made by whoever is holding the spanner.

Acceptance criteria written down in advance. Clearances, runout limits, balance grades and vibration acceptance should be stated before the machine comes apart, referenced to the standard being worked to rather than to the contractor's judgement. ASME performance test codes and the ISO balance quality standards are the usual references.

A parts position. Which parts are in stock, which have lead times, which are being reverse-engineered, and what happens to the schedule if an unstocked part is needed. This is where repairs actually slip.

A test and handover definition. What gets tested, at what load, against what criteria, and what documentation arrives at the end. A repair with no acceptance test is a repair whose success is a matter of opinion.

The five parts of a repair scope worth signing. The absence of any one of them is a warning. Source: MIMAH engineering practice.

Lead Times and What Actually Delays You

Repair schedules rarely slip on labour. They slip on three things.

Parts. Blades, diaphragms and rotors are long-lead items, and lead times lengthen when the original manufacturer no longer supports the machine. The mitigation is knowing your parts position before the outage rather than discovering it during one. On older machines, reverse engineering is often faster than chasing a manufacturer route that has quietly closed.

Findings. The rotor that was expected to clean up and does not, the diaphragm crack that was not visible until the deposit came off. This is what hold points and contingency pricing exist for.

Balancing and machining capacity. Rotor work needs a balancing machine of the right capacity and a workshop that is not already full. On a plant outage that lands in a regional peak season, capacity can be the binding constraint rather than the engineering.

The practical response is to compress the unknowns before the outage starts. Stage pressure surveys, vibration data with phase, thrust position history, borescope inspection where access allows, and a parts review against the actual serial number. Every one of those moves work out of the critical path.

The three things that actually delay a turbine repair. None of them is labour, which is what most schedules are built around. Source: MIMAH engineering practice.

Questions to Ask a Repair Contractor

Six questions separate contractors quickly.

What data do you need from us before you can scope this? A contractor who needs nothing is not scoping.

What is fixed price and what is provisional? And what triggers the provisional work.

Where are the hold points, and who decides at each one?

What acceptance criteria will you work to, and against which standard?

Which parts do you hold, and what is the lead time on the ones you do not?

What documentation do we get at handover? Clearance records, examination reports, balance certificates and as-left readings are the evidence that the repair was done to the standard claimed.

The answers tell you whether you are buying engineering or labour. Both have their place, but only one of them fixes a machine whose fault has not yet been correctly identified.

Where Repair Fits in the Wider Programme

Repair is reactive by definition, and a plant that repairs often is telling you something about its maintenance programme rather than its turbines. The machines that need unplanned repair are usually the ones whose condition was not being watched: no vibration trend, no stage pressure survey, no oil analysis, no record of what the thrust position was doing over the last two years.

A turbine on a proper maintenance and overhaul programme still needs repair occasionally, because things break. The difference is that the repairs are smaller, they are scoped from data rather than from wreckage, and they happen inside planned outages rather than instead of production.

MIMAH's rotating equipment team scopes and executes turbine repair on single-stage and multi-stage machines, including reverse engineering where original manufacturer support has lapsed. If you have a machine with a fault you cannot pin down, or a repair quotation you want reviewed before you sign it, get in touch.