Case Study: Shutdown Survey at Victorian Power Station

TL;DR: This shutdown survey power station Victoria case study follows a Latrobe Valley brown-coal generator replacing a worn pulverising mill and three burner assemblies during a fixed 14-day unit outage. ISS laser scanned the boiler house and mill bay before demolition began, delivered a registered point cloud and clash-checked as-built model inside 72 hours, and confirmed the new mill foundation and ducting tie-ins fitted to millimetre tolerances before fabrication was committed. The replacement dropped into a verified envelope, the unit returned to the grid on schedule, and a A$240,000 fabrication rework risk was closed out for a A$16,500 survey fee.

Key takeaways

• A terrestrial laser scanning shutdown survey, captured in the first 36 hours of a 14-day unit outage with a FARO Focus Premium and a Leica RTC360, gave the project a millimetre-accurate as-built of the mill bay and boiler penetrations before any steel was cut — the only window in which that geometry could be recorded.
• The point cloud was registered to better than 3 mm and tied to a local control net on GDA2020/MGA2020 Zone 55 with levels on AHD, so the new mill baseplate, coal piping and primary-air ducting could be set out against the same datum the structure was actually built to, not against drawings last revised in the 1980s.
• Clash detection against the OEM replacement-mill model found a 41 mm interference between the new primary-air duct flange and an existing structural brace — a clash that, undetected, would have stopped the lift mid-outage and forced an emergency redesign.
• The work was assessed against AS 3990 for the mechanical steelwork tolerances and the OEM installation envelope rather than any single surveying standard, with all instruments carrying ISO/IEC 17025-traceable calibration.
• The full scope — pre-demolition scan, registered point cloud, scan-to-model and post-installation verification — was delivered for A$16,500 inside the existing outage, against a fabrication-rework and outage-extension exposure the project costed at more than A$240,000 per day of slip.

The plant and the problem

The client operates a brown-coal-fired generating unit in Victoria's Latrobe Valley, feeding the National Electricity Market. The asset at the centre of this case study is one of the unit's pulverising mills and its associated burner front: a vertical spindle mill grinding raw lignite, feeding pulverised fuel through coal piping to burner assemblies penetrating the boiler's furnace wall. Brown-coal plant of this vintage runs hard, and the mill in question was past its economic wear life — grinding elements worn, throat ring eroded, and vibration trending up.

The maintenance plan was set. During a fixed 14-day unit outage, the mechanical contractor would strip out the old mill, replace three burner assemblies, and install an OEM replacement mill with new coal piping and a re-worked primary-air duct. The replacement was being fabricated off site against the original installation drawings.

That was the risk. The original boiler and mill bay were commissioned decades ago, the steelwork had been modified across multiple past outages, and the as-built drawings were a patchwork of revisions — some hand-marked, some never updated. The project engineer had two questions. Does the plant, as it physically stands today, match the drawings the new mill is being built to? And can that be confirmed early enough to correct any fabrication before the unit is opened up and the clock starts on the outage?

Why the survey had to happen first

A power station shutdown is the most expensive and least forgiving environment in industrial surveying. The outage window is fixed, every trade is sequenced against every other, and a unit off the grid in a tight market is losing revenue every hour. There is no room to discover, on day eight, that a new duct does not fit.

The instinct is to trust the drawings. On a plant this age, that is exactly the wrong move. Forty-year-old structures move, get modified, and accumulate undocumented changes. The only reliable record of what is actually there is a measurement of what is actually there — and the only time to take it on the critical equipment is before demolition, while the geometry the new plant must mate to still exists.

So ISS mobilised on the first day of the outage, ahead of the strip-out crews. As soon as the mill was isolated, locked out and made safe under the site's permit-to-work, the survey team scanned the mill bay, the surrounding boiler structure, the existing coal piping runs and the furnace-wall penetrations. Capturing this geometry first is what let the project compare reality against the fabrication drawings while there was still time to act on the difference.

The scan and finding the clash

ISS used a Leica RTC360 for the open structural volume of the mill bay and a FARO Focus Premium for the tighter approaches around the coal piping and duct flanges, where short-range detail mattered more than reach. Forty-one scan positions were tied together using cloud-to-cloud registration backed by black-and-white targets on the heavy steel, achieving a registered network better than 3 mm. The whole point cloud was constrained to a local control net surveyed with a Leica MS60 MultiStation on GDA2020/MGA2020 Zone 55, with levels carried from site AHD benchmarks.

The registered cloud was then taken into the office and processed scan-to-model: the existing structural steel, the mill foundation, the primary-air duct route and the furnace penetrations were modelled as solids and overlaid against the OEM's replacement-mill and new-duct geometry. This is where the survey earned its fee.

The clash check found a 41 mm interference between the new primary-air duct's lower flange and an existing structural brace that did not appear on the drawing the duct had been designed to. The brace had been added in a past outage and never drawn up. As designed, the new duct would have fouled the brace on the way in — a clash that would have surfaced only when the section was craned into position mid-outage, with the unit open and the fabricated steel already on site. The fix at that point would have been an emergency field re-cut and re-weld, or sending the section back to the workshop, either of which threatens the restart date.

Because the clash was found inside 72 hours, the duct flange detail was modified at the fabricator before the section shipped. The interference was designed out on paper, not discovered on the hook.

Equipment and method

All instruments carried current ISO/IEC 17025-traceable calibration certificates. The mechanical steelwork and baseplate tolerances were assessed against AS 3990 and the OEM installation envelope; the registered point cloud was delivered in E57 and recipe-built into the client's modelling environment as both a navigable cloud and a clash-ready solid model. Confined approaches inside the boiler structure and work near isolated plant were covered by the site's permit-to-work and ISS's own SWMS, with scanning sequenced around the demolition and rigging crews so no trade waited on the survey and the survey waited on no trade.

The result

With the clash designed out, the strip-out and fabrication proceeded against a verified envelope. When the new mill foundation was prepared, ISS set out the baseplate position and levels from the same control and AHD datum used for the scan, so the mill landed where the model said it would. After installation, a verification scan confirmed the as-installed mill, coal piping and primary-air duct against the design model before the unit was boxed up.

The mill replacement and burner work were completed within the 14-day outage, and the unit was returned to the grid on schedule with no extension. The new plant dropped into an envelope that had been measured rather than assumed, and the fabricated sections fitted on the first attempt.

The outcome for the operator

The commercial case is stark. The full ISS scope — pre-demolition scan, registered point cloud, scan-to-model clash detection and post-installation verification — came to A$16,500. The exposure it closed out was a mid-outage fabrication clash, which the project team costed at more than A$240,000 for a single day of outage slip once lost generation revenue, idle contractor labour across the whole shutdown, and emergency rework were included. Finding the 41 mm interference on a screen during week one, rather than on the crane hook in week two, was the difference between a planned line item and an unplanned blowout.

Beyond the avoided rework, the operator now holds a millimetre-accurate, datum-correct 3D record of the mill bay and boiler penetrations — the as-built that never properly existed. The next time this unit is opened for a mill, burner or duct change, the project starts from measured reality rather than a patchwork of marked-up drawings, and the scoping risk that drove this survey is permanently reduced.

Frequently asked questions

Why scan a power station before the shutdown rather than relying on the drawings?

Because plant of this age no longer matches its drawings. Decades of modifications, repairs and undocumented changes mean the only reliable record of what is physically there is a measurement of what is physically there. In this case the scan found a structural brace that was never drawn, which would have clashed with the new duct — a problem the drawings could never have revealed.

How accurate is a laser scanning shutdown survey on a structure this size?

The registered point cloud held better than 3 mm across 41 scan positions, with the underlying instruments accurate to roughly 2 mm at 10 m. Tied to a control net on GDA2020/MGA2020 Zone 55 and levelled to AHD, that is tight enough to set out a mill baseplate and clash-check fabricated steel with confidence. The limiting factor is registration and control discipline, not the scanner.

Can the scan really be turned around fast enough to act on it mid-outage?

Yes. The pre-demolition scan was captured in the first 36 hours and the registered cloud and clash model were delivered inside 72 hours — fast enough to re-detail the duct at the fabricator before the section shipped. Speed is the whole point of a shutdown survey; a result that arrives after the lift is worthless.

Which standards apply to a shutdown survey like this?

There is no single Australian Standard for industrial laser scanning, so the work is assessed against the relevant engineering standard for the task. Here the mechanical steelwork and baseplate tolerances were checked against AS 3990 and the OEM installation envelope, with all measurement on ISO/IEC 17025-traceable instruments and control on GDA2020/MGA2020 and AHD. Methodology and traceability are what make the result defensible.

Does the survey interfere with the demolition and installation crews?

No. Scanning is sequenced around the other trades — the pre-demolition scan goes in once the plant is isolated but before strip-out, and verification scanning fits between installation milestones. On this outage the survey neither waited on another trade nor held one up, which is essential in a fixed power station window.

Industrial Spatial Solutions delivers shutdown, turnaround and outage surveys for power generation, mineral-processing and heavy-industrial operators across Victoria and nationally, using Leica and FARO laser scanners, total stations and traceable control to turn ageing, poorly documented plant into millimetre-accurate as-built models before the steel is cut. If you have a unit outage on the calendar and a replacement or tie-in that has to fit first time, call ISS on 0407 057 015 to scope the survey around your shutdown window and request a fixed-price quote.