Control Network Methods Comparison

TL;DR: This control network methods comparison weighs the four ways Australian surveyors establish project control — static GNSS, RTK GNSS, total station traverse, and precise levelling — against the accuracy, terrain and budget realities of mining and industrial sites. No single method wins outright: a Pilbara iron ore expansion and a SAG mill alignment in Kalgoorlie need very different networks. The right answer is almost always a hybrid, anchored to GDA2020/MGA2020 and AHD, and chosen against the tightest tolerance the site has to hold.

Key takeaways

• Static GNSS delivers the strongest horizontal control over distance — 3-5 mm + 0.5 ppm baseline accuracy with a Leica GS18 or Trimble R12i pair — but its vertical component (10-20 mm typical) is too weak for tight levelness work and must be supplemented by levelling.
• RTK GNSS is the workhorse for densifying tertiary control and set-out at 8-15 mm horizontal / 15-30 mm vertical, but it cannot establish primary control to ICSM Second Order on its own and fails in pit voids, under conveyor galleries, and beside steel structures.
• Total station traverse (Leica TS60, Trimble S9) remains the only practical primary method in GPS-denied environments — underground, inside process buildings, in deep cuttings — reaching 1-5 mm relative over short braced figures.
• Precise levelling with a digital level and invar staff (Leica LS15, Trimble DiNi) is non-negotiable wherever vertical tolerance drops below ~10 mm: kiln piers, crane rails, mill soleplates, and AHD height transfer.
• For a control network methods comparison to be honest it must price the whole programme: establishing control is typically 5-10% of survey cost, but a failed datum can invalidate an entire shutdown's worth of measurements at $15,000-$50,000 per day of lost production.

What "control network method" actually means

A control network is the framework of permanently marked points — with known eastings, northings and reduced levels — that every other measurement on a site connects to. The "method" is how you determine those coordinates and tie them to a national datum.

In Australia that datum is GDA2020, projected to MGA2020 grid coordinates, with heights on the Australian Height Datum (AHD). Networks are classified under ICSM SP1 (Standards for the Australian Survey Control Network), which sets accuracy by order — from Zero Order (~1 mm relative, for deformation monitoring) down to Third Order (~50 mm, for general earthworks).

The methods compared here are not competitors so much as tools with different jobs. The skill is matching the method to the tolerance, the terrain, and the time you have on site — which is exactly what this control network methods comparison sets out to do.

The four core methods, head to head

Static GNSS

Two or more survey-grade receivers (Leica GS18 T, Trimble R12i) log satellite observations simultaneously for 20 minutes to several hours, and the baselines between them are post-processed against known marks or the AUSPOS/CORS network.

• Strengths: Excellent over long baselines; no line-of-sight needed; ties directly to GDA2020 via Geoscience Australia CORS. Ideal for primary control spread across a Pilbara or Bowen Basin lease where points sit kilometres apart.
• Weaknesses: Slow per point; weak height component; needs open sky, so useless inside buildings or against high pit walls.
• Typical accuracy: 3-5 mm + 0.5 ppm horizontal; 10-20 mm vertical.

RTK GNSS

A base receiver broadcasts corrections to a rover in real time (or the rover uses a network RTK/NTRIP service), giving instant coordinates as the surveyor walks the site.

• Strengths: Fast densification of secondary and tertiary control; ideal for set-out pegs, stockpile bases, drone ground control points (GCPs), and haul-road work. One operator can place dozens of marks a day.
• Weaknesses: Multipath near steel and concrete corrupts fixes; canyon effects in deep pits; not robust enough for ICSM Second Order primary control. Always check into a known mark before and after.
• Typical accuracy: 8-15 mm horizontal; 15-30 mm vertical.

Total station traverse

An instrument (Leica TS60, Trimble S9 HP) measures angles and distances between intervisible marks, run as a closed or braced traverse so misclosure can be detected and distributed by least squares.

• Strengths: Works where GNSS cannot — underground, inside an alumina refinery, in a box cut. Outstanding relative accuracy over short distances; the backbone of mechanical and dimensional control.
• Weaknesses: Needs intervisibility and stable setups; error accumulates over long open traverses; slower over large open ground than GNSS.
• Typical accuracy: 1-5 mm relative over a well-braced figure.

Precise levelling

A digital level (Leica LS15, Trimble DiNi 0.3) reads a barcoded invar staff to transfer height between benchmarks in closed loops.

• Strengths: The only method that reliably delivers sub-millimetre to few-millimetre height accuracy. Essential for AHD transfer, crane rail levels, mill soleplate flatness, and settlement monitoring.
• Weaknesses: Height only — no horizontal position. Slow over long distances and broken terrain.
• Typical accuracy: 0.5-2 mm per kilometre double-run.

Decision table: which method for which job

Read across the row that matches your tightest tolerance, not your average one. A plant set-out that is mostly ±25 mm but includes a single ±2 mm machine soleplate is, for control purposes, a ±2 mm job.

How the methods stack up on the factors that matter

Accuracy

For horizontal position over distance, static GNSS leads. For relative accuracy over short spans — the regime mechanical surveys live in — total station traverse wins. For height, precise levelling beats everything by an order of magnitude. RTK is the convenience option, not the accuracy option, and it should never carry primary control alone.

Speed and site disruption

RTK is fastest for covering ground; static GNSS is slow but unattended once logging starts; total station is steady but needs intervisibility cleared; levelling is the slowest per kilometre. On a live shutdown where every hour costs production, the method mix is often dictated by how fast control can be stood up around plant that is still partly running.

Terrain and environment

This is where the comparison gets decisive. Deep pits, conveyor galleries, process buildings and underground drives are GNSS-hostile — multipath and sky obstruction kill the fix. In those zones total station traverse is not the better choice, it is the only choice. Conversely, a flat, open Pilbara lease is exactly where GNSS shines and total station traversing would waste days.

Cost

Establishing control is typically 5-10% of total survey cost. GNSS reduces field hours on open sites; total station and levelling are labour-intensive but irreplaceable for tight work. The real cost driver is not the method's day rate — it is the cost of getting the datum wrong. A control failure discovered mid-shutdown can strand every downstream measurement, with downtime running $15,000-$50,000 per day on a major plant.

Why the answer is almost always hybrid

On a real Australian industrial site the methods combine rather than compete:

1. Static GNSS establishes primary control tied to GDA2020/MGA2020, spread across the lease and connected to the nearest CORS or AUSPOS solution.
2. Precise levelling transfers accurate AHD heights onto the primary marks, fixing the weak GNSS vertical.
3. Total station traverse densifies into GNSS-denied zones — inside the plant, down the decline, against the pit wall — carrying the primary coordinates with it.
4. RTK GNSS fills in tertiary set-out, GCPs and volume work off the adjusted secondary network.

The whole network is then adjusted by least squares so misclosures are detected and redundancy proves the result. That adjustment, not the brand of receiver, is what makes the control defensible.

Frequently asked questions

Can I run a whole mine control network on RTK GNSS alone?

No. RTK is excellent for set-out, stockpile work and GCPs, but at 8-15 mm horizontal and 15-30 mm vertical it does not meet ICSM Second Order for primary control, and it is unreliable near steel, in pit voids and under structures. Use static GNSS for primary marks and reserve RTK for densification — and always check RTK work into a known mark.

How accurate does control need to be for mill or kiln alignment?

Mechanical alignment lives in the 1-5 mm regime, which means a braced total station frame plus precise levelling, effectively First to Zero Order locally. General GNSS control is not tight enough to carry alignment work; the local equipment frame is established by traverse and levelling and only loosely tied to site grid.

Do I always have to connect to GDA2020 and AHD?

For anything that integrates with mapping, GIS, statutory plans, neighbouring leases or long-term monitoring, yes — connect to GDA2020/MGA2020 and AHD. A purely local frame is acceptable only for isolated, short-lived equipment work where no external integration is required. Even then, recording the transformation to GDA2020 protects you later.

What about drone surveys — do they need ground control?

Yes. UAV photogrammetry under CASA Part 101 still relies on surveyed ground control points (or PPK with a logged base) to anchor the model. Those GCPs are placed by RTK or total station off your established control network. The drone does not replace control — it consumes it.

How often should the control network be re-checked?

It depends on the environment. Blasting, ground movement, settlement and construction traffic all degrade marks. Mine primary control is typically re-observed quarterly; deformation control as often as weekly; permanent industrial control at least annually. Always re-verify before a major shutdown rather than trusting marks last checked months ago.

Talk to a surveyor before you pick a method

Choosing control methods from a table is a starting point, not a specification. The right network depends on your tightest tolerance, your terrain, your datum requirements and your shutdown window — and getting it wrong is one of the few survey errors that can invalidate everything built on top of it.

Industrial Spatial Solutions designs, establishes and adjusts control networks to ICSM standards across mining, processing and construction sites Australia-wide, combining static and RTK GNSS, total station traverse, precise levelling and laser scanning into a single defensible framework on GDA2020/MGA2020 and AHD. Call us on 0407 057 015 to discuss your site and get a fixed-price quote for control that holds.

Industrial Spatial Solutions — the right method, the right datum, control that holds.

Related reading: Control network surveys explained, What is dimensional control, Engineering and civil surveys