Case Study: Control Network at WA Gold Mine

TL;DR

This control network gold mine WA case study follows an open-pit operation in the Eastern Goldfields where dig faces, dump designs and grade-control models had stopped agreeing with one another. ISS found the site's primary control network had drifted and degraded since establishment, with two pillars physically disturbed and the rest never properly adjusted as a single network. ISS re-observed a 14-station network with Leica GS18 GNSS and an MS60 MultiStation, performed a rigorous least-squares adjustment on GDA2020/MGA2020 Zone 51 with AHD heights, and delivered a homogeneous network good to better than 10 mm horizontal — restoring agreement between survey, mine planning and grade control across the whole lease.

Key takeaways

• Conflicting bench levels, mismatched dump volumes and grade-control blocks landing in the wrong place all traced back to one root cause: a primary control network that had never been adjusted as a single, statistically validated whole.
• Two of the original eight pit-perimeter control marks had moved — one undercut by a cut-back, one heaved on reactive clay — yet they were still being occupied daily, quietly contaminating every setout and pickup that referenced them.
• ISS re-observed 14 stations as one network using Leica GS18 GNSS (static and RTK) tied to three AUSCORS/AUSPOS-validated base points, with an MS60 total station bridging the pit floor where GNSS sky view was poor.
• A least-squares adjustment in GDA2020/MGA2020 Zone 51 (AHD via AUSGeoid2020) returned 95% confidence semi-major axes under 9 mm horizontal and 14 mm vertical — a network the mine could trust for grade control, volumes and statutory survey.
• Field work took two days; the adjusted coordinate register, error ellipses and a marked-up disturbed-mark report were issued within five business days. Total fee was AUD 18,600 including mobilisation from Kalgoorlie.

The challenge: when nothing on site agrees

The client runs a mid-tier open-pit gold operation in the Eastern Goldfields region of Western Australia, the country's gold heartland around Kalgoorlie, Leonora and Laverton. The pit had deepened through several cut-backs, the waste dumps had grown, and a satellite pit had been opened on the same mining lease. Through all of that, the site relied on a primary control network established years earlier when the operation was a fraction of its current size.

The symptoms arrived from every direction. The mine surveyors found bench toe and crest levels that disagreed with the design by 30-60 mm depending on which control mark they set up over. Monthly dump volumes reconciled poorly against truck-count and mill-feed figures. Most seriously for a gold operation, grade-control drilling and the ore-block models built from it were starting to misregister against the dig faces — dig blocks were being mined slightly off-position, the kind of error that sends ore to the waste dump and dilution to the mill.

The site's first instinct was that the GNSS rovers or the machine-guidance system were at fault. They were not. When the same point was checked from two different control marks and gave two different answers, the problem was not the instruments — it was the control they were all hanging off. A network that is internally inconsistent makes every downstream measurement inconsistent in a way no amount of careful field technique can correct.

The approach: rebuild the network as one adjusted whole

ISS scoped this as a primary control re-establishment and rigorous network adjustment, not a quick re-peg. The brief was to deliver a single, homogeneous, statistically defensible control network on the national datum that every team — survey, geology, mine planning, machine guidance and the statutory mine surveyor — could share with confidence.

The decision that mattered most was made before any observation: the new network would be solved as one least-squares adjustment, not built up piecemeal by radiating from one "good" mark to the next. The original network had failed precisely because it had grown that way — marks added over the years by traverse and resection off whatever was nearby, never tied together and never adjusted as a system. Errors had accumulated invisibly. A whole-network adjustment, by contrast, distributes residuals across all observations and produces a defensible error estimate for every station, so the site finally knows not just where each mark is but how well it is known.

The work was anchored to GDA2020 and projected to MGA2020 Zone 51, the correct zone for this part of WA, with AHD heights derived through the AUSGeoid2020 model. Three of the most stable existing marks, sited on competent rock well clear of mining activity, were re-validated by long-occupation static GNSS post-processed through Geoscience Australia's AUSPOS service before they were trusted as the network's foundation.

ISS mobilised a two-person crew from its Kalgoorlie base, which kept mobilisation short and let the field programme start the morning after arrival.

Equipment and method

The campaign combined long static GNSS sessions on the foundation marks with shorter sessions and redundant RTK baselines across the perimeter marks, so that every station was observed from more than one independent setup — redundancy is what makes a least-squares adjustment meaningful. Where the pit floor sat below the rim and GNSS sky view was blocked by high walls, the MS60 carried precise angle-and-distance observations down from rim control, tying the floor marks into the same adjustment rather than leaving them on a separate, weaker footing. Every observation went into a single adjustment so the entire lease — main pit, satellite pit and dumps — sat in one consistent frame.

What the survey found

The adjustment told a clear story. Before any new observations were even processed, the redundant measurements exposed two control marks that simply did not fit the rest of the network within tolerance — they were flagged as outliers and pulled from the foundation set.

Both disturbed marks — PIT-03 undercut by an advancing cut-back, PIT-06 heaved on reactive clay beside a haul road — were still in daily use. Every setup over PIT-03 had been quietly throwing a ~40 mm horizontal error into whatever it touched, which neatly explained the off-position dig blocks on that side of the pit. The legacy network as a whole showed an internal spread of up to roughly 55 mm once forced into a common adjustment, confirming that the real problem was systemic inconsistency, not a single bad mark.

After the full least-squares adjustment of the rebuilt 14-station network, the picture inverted: a tight, homogeneous frame with realistic, small error estimates on every point.

This is the difference dense, redundant observation and a rigorous adjustment make: the site went from a network where nobody could say how good any mark was, to one where every station carries a defensible sub-10 mm horizontal error estimate.

The result: one frame the whole site shares

With the adjusted coordinate register loaded, the downstream conflicts resolved. The mine surveyors re-checked the bench levels that had disagreed with design and found them landing within a few millimetres once set up over the validated marks. The grade-control geologists re-registered the recent drill collars into the corrected frame, which pulled the ore-block models back into alignment with the dig faces — the off-position blocks were a control problem, not a geology problem.

ISS delivered the adjusted coordinate register, the full error-ellipse output for every station, and a marked-up site plan identifying the decommissioned marks and the new replacement marks, with witness sketches and recovery notes for each. The machine-guidance and RTK base configurations were updated to the new coordinates so dozers, drills and the survey rovers all referenced the same network from day one.

The outcome: a defensible baseline, not a guess

The control network re-establishment cost AUD 18,600, including mobilisation from Kalgoorlie, two days of field observation, the network adjustment and reporting, and AUSPOS validation of the foundation marks. Against that, the operation had been losing ore to dilution and misplaced dig blocks — a single misregistered grade-control panel can be worth far more than the survey in lost gold ounces and misrouted material, before the wasted dump-volume reconciliation effort is even counted.

Just as important, the site now has a network whose accuracy is documented rather than assumed. ISS recommended bringing the primary control into a routine re-validation cycle — annual re-observation of the foundation marks and a check of perimeter marks after each major cut-back, since active mining will always disturb marks near the crest. The client adopted the cycle and added the foundation marks to a protected-survey-infrastructure register so they are not built over or undercut again without warning.

Frequently asked questions

Why did the GNSS rovers seem to be the problem when they were not?

A GNSS rover is only ever as good as the control it is initialised against. When the same ground point checked from two control marks gave two different coordinates, the rovers were working correctly — they were faithfully reporting positions relative to a network that was internally inconsistent. Fixing the rovers or the base would have changed nothing; the fault was in the control, which is why a full network re-establishment and adjustment was the right call rather than re-pegging individual marks.

Why does the network have to be adjusted as a single least-squares solution?

Building control by radiating from one mark to the next lets small errors accumulate without ever being detected, which is exactly how the legacy network degraded. A rigorous least-squares adjustment uses the redundant observations to distribute residuals across the whole network and, critically, produces a realistic error estimate for every station. That is the difference between knowing where a mark is and knowing how well you know it — and on a gold operation, that confidence is what protects grade control and volumes.

Why GDA2020 / MGA2020 Zone 51 and AHD for a mine site?

GDA2020 is Australia's current national datum, projected to MGA2020 Zone 51 for the Eastern Goldfields, with AHD heights derived through the AUSGeoid2020 model. Putting the whole lease — main pit, satellite pit and dumps — on the national datum means survey, geology, planning and machine guidance all share one consistent frame, and the data ties cleanly to tenement, environmental and statutory mine-survey requirements rather than living in an undefined local grid.

How were the disturbed control marks detected?

Redundancy. Because every station was observed from more than one independent setup, the adjustment could test each mark against the others. PIT-03 and PIT-06 failed those statistical tests and were flagged as outliers before being confirmed on the ground — one undercut by a cut-back, one heaved on reactive clay. A network with no redundant observations cannot detect a disturbed mark at all, which is how they stayed in daily use.

What does a control network re-establishment like this cost?

This job was AUD 18,600 including Kalgoorlie mobilisation, two days of field observation, AUSPOS validation, the least-squares adjustment and reporting. Most primary control re-establishments on a WA mine site fall between AUD 10,000 and AUD 25,000, driven by network size, the number of marks needing replacement, pit access and GNSS sky-view conditions. Establishing the work inside an existing site presence and protecting the foundation marks afterward keeps recurring re-validation costs much lower.

Request a quote

If bench levels, dump volumes and grade-control blocks have stopped agreeing on your site, the cause is very often a primary control network that has drifted or was never adjusted as a single whole — and it is far cheaper to fix than to keep losing ore to misplaced dig blocks. ISS re-establishes and rigorously adjusts mine control networks across the Goldfields and nationally, using Leica GNSS and total stations on GDA2020/MGA2020 with AHD heights, and delivers a defensible coordinate register with documented accuracy for every mark. Call us on 0407 057 015 to scope a control network re-establishment or to set up an annual re-validation cycle before your next cut-back.