News & Articles
.jpg)
VARNISH: Closing the Loop: Economics of Adaptive Varnish Control in Mobile Mining Equipment (Part 3)
Article by William Gillette (LogiLube, LLC)
The business case for varnish control rarely begins with the hydraulic oil.
It begins with the machine.
Introduction
In Part 3 of this Varnish Series, the discussion moves from the chemistry and mechanisms of varnish formation to the practical and economic application of varnish control on ultra-class mobile mining equipment. Using the Liebherr R9800 hydraulic excavator operating in the Pilbara iron-ore environment as the primary case study, this article examines how varnish, oxidation, contamination and fluid degradation can affect hydraulic performance, loading cycles, equipment availability, tonnes moved and ultimately EBITDA (earnings before income tax and amortization). It compares the conventional practice of fixed-interval hydraulic-oil replacement with a ‘condition-based’ strategy built around SmartOil G3™ Autonomous Fluid Intelligence™, multi-domain sensing, Exception Sampling™, laboratory confirmation and Adaptive Dosing™ of Fluitec DECON™. The article also explores the economics of extending hydraulic-oil service life, the limitations of a complete reservoir drain, the value of maintaining an approved DECON treatment concentration, and the potential fleet-wide impact when this approach is applied across Fortescue's large hydraulic-excavator population. The objective is to demonstrate how fluid condition management can evolve from a maintenance activity into a measurable strategy for protecting productive hours, reducing unnecessary interventions and optimizing mining EBITDA.
Background
Varnish has long been recognized as a reliability issue in fixed industrial assets such as paper machines, steam turbines and gas turbines, where deposits can affect servo valves, lubrication systems, heat exchangers and other precision components. Those applications demonstrate the technical consequences of fluid degradation.
Mobile mining equipment adds another dimension: production economics.
In a large Pilbara iron-ore operation, an ultra-class hydraulic excavator is not simply an equipment asset. It is a production node. Its role is to continuously load a fleet of haul trucks feeding the mine’s material-handling and processing system.
When the excavator slows, haul-truck productivity can decline.
When the excavator stops, a significant portion of the production chain may stop with it.
That changes the economic meaning of hydraulic-oil varnish.
A hydraulic valve that becomes progressively sluggish, a cooler that loses heat-transfer efficiency or a hydraulic circuit that begins operating several degrees hotter than normal may initially appear to be maintenance issues.
On a machine such as the Liebherr R9800, however, those conditions can ultimately become:
Longer loading cycles → truck queueing → reduced tonnes moved → lower equipment availability → lost production and EBITDA.
By the time varnish becomes visible as a maintenance problem, the hydraulic fluid may have been moving toward failure for hundreds or thousands of operating hours.
.jpg)
Adaptive Dosing™
SmartOil G3™ Adaptive Dosing™ is intended to intervene earlier in the varnish-degradation cycle by linking fluid condition directly to machine condition and, ultimately, to production economics. Rather than relying on a fixed calendar interval, a manual additive treatment or a single sensor threshold, Adaptive Dosing™ uses continuously updated fluid-condition and machine-operating data to determine whether treatment is required, when it should occur and how much treatment chemistry should be delivered.
The system continuously monitors the circulating hydraulic fluid using a multi-domain sensor architecture and evaluates changes through the G3 Edge-AI Brain™. Inputs can include viscosity, density, dielectric constant, Tan Delta/EIS response, particle contamination, water, hydraulic temperature, filter differential pressure, reservoir level, operating hours and other machine-state information. The Edge-AI Brain™ evaluates not only individual measurements, but also their rate of change, correlation with other sensor domains, temperature-normalized behavior and deviation from the machine's established healthy-fluid baseline.
When the combined evidence indicates an abnormal condition, Exception Sampling™ can automatically capture a representative physical oil sample while the condition is occurring. Laboratory analysis can then confirm parameters that cannot yet be measured directly and reliably on the machine, including MPC varnish potential, antioxidant reserve using RULER, FTIR oxidation, TAN, Karl Fischer water, ISO 4406 particle count, PQ index and elemental analysis.
When the evidence confirms that varnish mitigation is appropriate, the G3 DOSE™ module provides the actuation layer for Adaptive Dosing™. Rather than making a large one-time manual chemical addition, the system can deliver a precisely controlled quantity of an approved treatment chemistry such as Fluitec DECON™, while the G3 Edge-AI Brain™ maintains a running fluid mass balance that accounts for hydraulic-system inventory, oil makeup, leakage, partial drains, reservoir drain-and-refill events and previous treatment additions. This allows the system to maintain an approved DECON concentration within the validated 3–5 vol% treatment range, subject to Fluitec recommendations, lubricant compatibility and site/OEM engineering approval.
Adaptive Dosing™ therefore operates as a bounded closed-loop control process. Treatment is not initiated simply because one sensor crosses one numerical threshold, nor is chemistry added merely because a predetermined number of operating hours has elapsed. The system combines multi-domain evidence, determines whether laboratory confirmation is warranted, calculates the appropriate treatment quantity, records the intervention and then monitors the post-dose response to verify that the hydraulic fluid is moving back toward its approved performance envelope.
The control loop is:
Measure → Interpret → Correlate → Exception Sample™ → Confirm → Calculate → Dose → Circulate → Verify → Repeat only when necessary
.jpg)
“The next evolution in varnish mitigation is not simply detecting varnish or adding chemistry. It is knowing the condition of the oil, maintaining the correct treatment concentration and verifying that the machine and lubricant respond as intended.” William Gillette, Cheif Executive Officer, LogiLube, LLC
This functionality is a key element of the SmartOil G3™ intellectual-property strategy and is the subject of a patent-pending application with the United States Patent and Trademark Office (USPTO) covering aspects of condition-responsive fluid treatment and Adaptive Dosing™ control.
For a large mining excavator such as the R9800, the value of this architecture extends well beyond reducing lubricant consumption. By detecting deterioration earlier, maintaining the appropriate treatment concentration and avoiding unnecessary drain-and-refill events, Adaptive Dosing™ is intended to help preserve hydraulic performance, reduce maintenance intervention and protect excavator availability.
For mining, the business objective remains equally straightforward:
Protect the productive hour.
DECON™ + SmartOil G3™: Closed-Loop Varnish Mitigation
Detecting a developing varnish condition is only part of the solution.
Once fluid analysis confirms that the R9800 hydraulic oil is developing elevated varnish potential—and once mechanical causes such as excessive temperature, contamination, cooling-system deterioration or abnormal component wear have been investigated—the next question is:
How should the fluid be treated?
For this application, Fluitec's DECON™ fluid-enhancement technology, and specifically DECON AW for hydraulic-oil applications, provides the treatment chemistry around which SmartOil G3™ Adaptive Dosing™ can operate.
Fluitec describes DECON as a lubricant-compatible varnish and deposit-control technology based on its patented Solvancer® chemistry. Rather than relying solely on filtration to remove insoluble contamination, Solvancer increases the ability of the circulating lubricant to keep varnish-forming degradation products in solution. Existing carbonaceous deposits can progressively dissolve back into the oil while the chemistry also reduces the tendency for additional deposits to form.
“Varnish management should not be viewed as a one-time cleanup event. The objective is to manage lubricant solvency and degradation continuously so deposits do not have the opportunity to control machine reliability.” Greg Livingstone, Chief Innovation Officer, Fluitec
For hydraulic systems, Fluitec identifies DECON AW as the formulation intended for applications where varnish, deposit formation and antiwear performance are concerns. Fluitec recommends a typical in-service treat rate of approximately:
3–5% by volume.
This concentration range creates a natural application for SmartOil G3™ Adaptive Dosing™.
From Manual Treatment to Controlled Concentration
A conventional DECON treatment might involve calculating a quantity of product, manually adding it to the reservoir and periodically checking the oil afterward.
That approach becomes more difficult on a mobile mining machine because the hydraulic-fluid inventory is continuously changing.
An R9800 may experience:
-
hose failures;
-
cylinder repairs;
-
oil leakage;
-
reservoir top-ups;
-
filter changes;
-
partial drains;
-
major maintenance events; and
-
complete reservoir drain-and-refill events.
Every litre of hydraulic oil added to the machine changes the concentration of DECON circulating in the system.
If 500 L of untreated hydraulic oil is added following a maintenance event, for example, the DECON concentration is diluted.
If 5,800 L is drained from the reservoir during a major service, a portion of the treatment chemistry leaves with the drained oil while another portion remains distributed through the residual oil trapped in pumps, valves, cylinders, motors, coolers, hoses and piping.
The target therefore cannot be managed reliably by simply asking:
"When was DECON last added?"
The more useful question is:
"What is the estimated DECON concentration in the hydraulic system right now?"
That is the role of SmartOil G3™ Adaptive Dosing™.
Treat the 10,000-L Hydraulic System as a Fluid Mass Balance
The R9800 contains approximately 10,000 L of hydraulic fluid across the complete hydraulic system, including approximately 5,800 L in the reservoir.
If Fluitec and the mine establish a nominal DECON AW treatment concentration within the recommended 3–5 vol% range, the corresponding theoretical treatment inventory for a fully mixed 10,000-L hydraulic system is approximately:
These values illustrate the total-system concentration concept. The actual commissioning quantity and approved operating target should be established with Fluitec, the hydraulic-oil supplier, Liebherr/site engineering and the mine's reliability organization before autonomous dosing is enabled.
Importantly, SmartOil G3 does not need to inject hundreds of litres at once.
The G3 DOSE™ architecture is intended to maintain the validated concentration through controlled incremental additions.
The Edge-AI Brain™ maintains an estimated fluid mass balance:
**Current hydraulic-oil inventory
-
new-oil additions
− known oil losses
− oil removed during maintenance -
previous DECON additions
− estimated DECON removed with lost or drained oil
= estimated current DECON concentration.**
“The chemistry becomes much more powerful when it is connected to accurate fluid inventory, condition data and controlled treatment. Maintaining the correct concentration as the system loses and receives oil turns deposit control into an active fluid-management process.” Dr. Cristian Soto, Chief Operations Officer, Fluitec
The system can then determine the small amount of DECON required to restore the approved concentration target.
This transforms DECON from a periodic maintenance treatment into a continuously managed fluid-performance variable.
Maintaining a Bounded 3–5% Operating Envelope
SmartOil G3 should not simply command DECON whenever a sensor value rises.
Nor should it continuously chase a fixed concentration without understanding what is happening to the machine.
A more robust control strategy establishes an approved DECON operating envelope.
For example, during engineering commissioning the mine and Fluitec might establish a nominal target concentration within the 3–5 vol% recommended range.
The G3 Edge-AI Brain™ then tracks:
-
estimated total hydraulic-fluid inventory;
-
hydraulic-oil makeup;
-
leakage;
-
partial drain events;
-
complete reservoir drains;
-
DECON additions;
-
operating hours;
-
hydraulic temperature;
-
viscosity;
-
dielectric behavior;
-
Tan Delta/EIS condition;
-
particle contamination;
-
water;
-
MPC laboratory results;
-
antioxidant condition; and
-
previous treatment response.
The G3 DOSE™ module can then deliver only the amount required to return the fluid to the approved treatment envelope.
Conceptually:
Measure → Calculate concentration → Verify need → Micro-dose → Circulate → Sample → Confirm response.
This is significantly different from conventional additive injection.
The treatment is controlled by both concentration and condition.
Condition Determines Whether Dosing Is Appropriate
Maintaining DECON within an approved concentration range does not mean that every abnormal fluid condition should initiate a dose.
An increasing MPC result accompanied by antioxidant depletion and otherwise normal machine operation may support additional treatment.
But elevated hydraulic temperature combined with normal varnish potential may indicate a cooler, fan or mechanical problem.
Increasing particle count may indicate dirt ingress or component wear.
Increasing water may require contamination removal rather than DECON.
A sudden viscosity change may indicate incorrect makeup oil or another contaminant.
In each case, adding more treatment chemistry could obscure the real problem if the underlying failure mode is not understood.
SmartOil G3 therefore uses its multi-domain sensor-fusion architecture to determine whether the observed condition is consistent with lubricant degradation before treatment is considered.
When uncertainty exists, the appropriate response is not automatically:
DOSE.
It is:
EXCEPTION SAMPLE™.
The physical sample can then be analyzed using MPC, RULER, FTIR oxidation, TAN, viscosity, Karl Fischer water, ISO 4406 particle count, PQ and elemental analysis.
Laboratory evidence determines whether the intervention should be:
Continue → Filter → Dose → Repair → Drain.
DECON as Part of the Condition-Based Drain Strategy
The combination of SmartOil G3 and DECON has broader implications for the R9800 hydraulic-oil drain strategy.
The traditional maintenance model attempts to manage degradation risk by periodically discarding part of the oil inventory.
The SmartOil G3 + DECON model attempts to manage the condition of the oil while it remains in service.
Fluitec has demonstrated this principle in hydraulic applications. In one published hydraulic-system case study, approximately 3% DECON was added to the circulating oil. Fluitec reported that varnish potential subsequently declined, contaminated components began cleaning and hydraulic performance improved. The operator continued monitoring MPC and planned to maintain the DECON concentration as new oil was added to the system.
That operating philosophy is highly relevant to mobile mining.
Instead of allowing degradation products to progressively accumulate until a 1,000-hour drain attempts to dilute them, SmartOil G3 can continuously observe the condition of the oil while DECON maintains an environment intended to keep varnish-forming material in solution.
The resulting maintenance model becomes:
Monitor continuously.
Capture exceptions.
Confirm with the laboratory.
Maintain the approved DECON concentration.
Correct mechanical causes of degradation.
Drain the oil when condition—not simply operating hours—justifies the intervention.
Adaptive Dosing™ Closes the Loop
Fluitec provides the chemistry.
SmartOil G3 provides the intelligence, fluid inventory model, physical sampling and dosing control required to apply that chemistry as part of a condition-based reliability strategy.
The combination creates a closed-loop architecture:
G3 multi-domain sensors
↓
G3 Edge-AI Brain™
↓
Exception Sampling™
↓
Laboratory confirmation: MPC + RULER + FTIR + TAN + supporting analysis
↓
Treatment decision
↓
G3 DOSE™ — controlled DECON AW addition
↓
3–5 vol% approved treatment envelope
↓
Continuous circulation through the R9800 hydraulic system
↓
Post-treatment verification
The objective is not simply to put an additive into the reservoir.
The objective is to accurately manage the chemistry required to keep the hydraulic fluid—and therefore the hydraulic system—inside an approved performance envelope.
For an R9800 whose productive operating hour can represent hundreds of thousands of dollars of iron-ore throughput, that distinction is economically important.
DECON provides the varnish-control chemistry.
SmartOil G3™ Adaptive Dosing™ determines when, why and how much should be delivered.
Autonomous Fluid Intelligence™: From Fluid Data to Production Protection
The operating principle behind SmartOil G3™ Adaptive Dosing is Autonomous Fluid Intelligence™—the ability to continuously observe the condition of the fluid, interpret what changing measurements mean for the machine and initiate an appropriate response without waiting for the next scheduled maintenance event.
Traditional oil analysis, machine telemetry, laboratory testing and additive treatment are usually independent activities.
SmartOil G3 combines them into a coordinated reliability process.
The G3 Edge-AI Brain™ can correlate real-time fluid measurements with hydraulic temperature, machine load, operating hours, reservoir level, oil additions, fluid losses and previous treatment history.
When the system recognizes an abnormal condition, Exception Sampling™ can capture a representative physical oil sample while the abnormal condition is actually occurring.
Laboratory analysis can then be used to confirm varnish potential, antioxidant depletion and other fluid-health indicators.
When treatment is justified, the G3 DOSE™ module can deliver a bounded micro-dose of the approved treatment chemistry. The subsequent fluid response is then monitored to establish whether the fluid is returning toward its approved operating envelope.
The objective is not simply to automate an additive pump.
The objective is to determine:
Is treatment actually required?
What condition caused the change?
How much treatment is appropriate?
Did the fluid respond?
Did machine performance respond?
That is the distinction between automated dosing and Autonomous Fluid Intelligence™.
Why Mobile Mining Changes the Varnish Equation
Hydraulic excavators operate under severe loading, shock, dust, thermal cycling and changing ambient conditions.
Their hydraulic reservoirs may also be repeatedly replenished because of hose replacement, cylinder maintenance, leakage or other repairs.
This creates a moving chemical target.
Makeup oil dilutes degradation products.
It replenishes some additive chemistry.
It changes the concentration of any previously applied varnish treatment.
It can also temporarily make a laboratory sample appear healthier without eliminating the underlying mechanism causing accelerated oil degradation.
A restricted cooler, for example, may increase hydraulic temperature and accelerate oxidation.
A fan-control problem may produce a similar effect.
Repeated leakage and makeup oil may partially dilute the resulting degradation products.
A conventional periodic sample sees only the condition of the fluid at the moment the sample was collected.
SmartOil G3 is intended to understand the history leading to that condition.
That capability becomes especially significant on an ultra-class excavator.
The Liebherr R9800 has a published hydraulic-system capacity of approximately 10,000 litres, including a hydraulic tank capacity of approximately 5,800 litres. Its attachment and travel systems use ten variable-flow pumps, each capable of approximately 750 L/min at pressures up to 320 bar.
This is approximately 2,642 gallons of hydraulic fluid supporting an 800-plus-tonne production machine.
The cost of the hydraulic oil is therefore only a small part of the economic equation.
The 1,000-Hour Hydraulic-Oil Drain Model
A representative maintenance practice encountered in severe-duty mobile mining applications is to perform a major hydraulic service at approximately 1,000 operating-hour intervals.
For purposes of this Pilbara economic model, Scenario A assumes that the 1,000-hour service includes:
-
draining the R9800 hydraulic reservoir;
-
replacing the drained hydraulic oil with new oil;
-
replacing the applicable hydraulic-filter elements;
-
inspection and service work;
-
returning the system to operating condition; and
-
completing the required post-maintenance checks before returning the excavator to production.
The R9800 operating documentation includes a scheduled 1,000-hour maintenance interval, although individual mines may apply different oil-drain practices depending on lubricant, contamination history, OEM guidance and site maintenance policy. Liebherr separately recommends regular used-oil analysis specifically as a means of optimizing hydraulic-oil change intervals.
The conventional approach provides a simple maintenance rule:
1,000 hours reached → drain oil → replace filters → refill system → restart machine.
The problem is that a reservoir drain is not equivalent to replacing all of the hydraulic fluid in the machine.
What Is the 1,000-Hour Oil Change Trying to Prevent?
The underlying purpose of a fixed 1,000-hour hydraulic-oil change interval is not to remove oil simply because it has accumulated 1,000 operating hours. It is to prevent the hydraulic fluid from progressing into a condition where contamination, degradation or loss of physical properties threatens the pumps, valves, cylinders and other precision components of the hydraulic system. Liebherr Mining documentation establishes several useful fluid-condition boundaries for large mining excavators. Hydraulic oil should normally be maintained at ISO 4406 cleanliness code 20/18/15 or cleaner and below 0.10% water by mass; in Liebherr condition-monitoring tables for large mining excavators, an ISO cleanliness result of 21/19/16 initiates corrective action, with filtration intended to restore the oil to 20/18/15 or better. Other published hydraulic-oil action levels include approximately 15% viscosity deviation from the new-oil baseline, 15 ppm silicon as an indicator of dirt ingress, and a PQ ferrous-particle index of 50, followed by filtration, investigation and resampling as appropriate. These limits address several major failure domains: particulate contamination that accelerates abrasive wear and valve sticking; water contamination that promotes corrosion, additive degradation and loss of lubricity; viscosity shift caused by oxidation, shear or contamination; airborne silica introduced through breathers, cylinders or maintenance activity; and abnormal ferrous debris associated with component distress. Varnish adds another failure domain that is not adequately represented by particle count alone. Oxidation products, antioxidant depletion and soluble or sub-micron varnish precursors can develop while conventional ISO cleanliness remains acceptable. I have not identified a publicly published Liebherr R9800 MPC or varnish-specific condemning limit; consequently, MPC, RULER antioxidant reserve, FTIR oxidation and related laboratory tests should be treated as supplementary condition indicators rather than represented as Liebherr OEM condemning criteria.
Multi-Domain G3 Sensing: Replacing the Hour Meter with Evidence
The opportunity for SmartOil G3™ is to replace the 1,000-hour clock as the primary proxy for oil condition with a multi-domain sensor architecture that continuously observes several independent properties of the hydraulic fluid. No single sensor should be expected to duplicate a laboratory oil analysis. Instead, multiple sensor domains should establish a continuously updated fluid-health fingerprint. The G3 Edge-AI Brain™ can compare that fingerprint with the machine’s own historical baseline, compensate measurements for oil temperature and operating state, identify abnormal rates of change and initiate an Exception Sample™ when the combined evidence indicates that laboratory confirmation is warranted.
The G3-SEN1™ is particularly valuable because the FPS2800 simultaneously measures viscosity, density, dielectric constant and temperature, allowing changes in several physical properties to be correlated rather than interpreting a single scalar measurement. TE specifically identifies hydraulic oils among the sensor’s applications and describes multi-parameter measurement as a means of improving fluid characterization. G3-SEN2™ adds an independent electrochemical condition domain; the OQSx-G2 continuously reports its Tan Delta Number and temperature and is intended to detect changing oil condition and contamination in real time. The OPCom II provides the complementary particle domain by directly reporting cleanliness according to ISO 4406 rather than attempting to infer particle contamination from another fluid property. An EIS channel such as the Poseidon Trident QW3100 can further strengthen the architecture by examining impedance across multiple frequencies while independently monitoring dissolved water, providing sensitivity to lubricant degradation, additive condition and contamination.
Exception Sampling™: The Used Oil Analysis Laboratory Becomes the Confirmation Layer
The objective of the sensor suite is not to eliminate laboratory oil analysis. It is to change when and why the laboratory sample is taken.
Instead of:
1,000 operating hours → drain oil because the hour meter reached the interval
the SmartOil G3 model becomes:
Continuously measure → identify abnormal multi-domain behavior → capture an Exception Sample™ → obtain laboratory confirmation → filter, treat, repair, continue operating or drain according to actual condition.
A practical G3 control model could establish an early-warning envelope before the Liebherr action boundary is reached. For example, viscosity deviation could generate a G3 warning as it approaches approximately 10% from the normalized new-oil baseline, leaving margin before an approximately 15% laboratory action level. Particle count could generate an escalating warning as the fluid trends toward ISO 21/19/16, while a sustained excursion to that level would automatically command an Exception Sample™. A rising water indication would similarly trigger sampling before the 0.10% OEM boundary is reached. These proposed G3 warning levels would be commissioning parameters—not substitute OEM condemning limits—and would be validated against the specific hydraulic oil, R9800 operating environment and laboratory history.
Most importantly for varnish, an Exception Sample™ should also be triggered when the combined dielectric, viscosity, EIS/Tan Delta and temperature-normalized degradation signature departs materially from the established healthy-oil fingerprint even though particle count, water and viscosity remain inside conventional limits.
That sample can then be submitted to the laboratory for a targeted confirmation slate:
MPC varnish potential; RULER antioxidant reserve; FTIR oxidation; TAN; kinematic viscosity at 40°C and 100°C; Karl Fischer water; ISO 4406 particle count; PQ index; and ICP elemental analysis including silicon and wear metals.
This creates a fundamentally different maintenance strategy.
The existing fixed-interval model asks:
“Has the R9800 reached 1,000 hours?”
Autonomous Fluid Intelligence™ asks:
“Is the hydraulic oil approaching a condition that actually requires intervention?”
That distinction is central to the economic argument for extending hydraulic-oil drain intervals. SmartOil G3 does not simply assume that oil can safely remain in service longer. It creates a continuous surveillance layer backed by autonomous physical sampling and laboratory evidence to demonstrate when continued operation is justified—and to identify developing contamination, varnish or component distress before the fluid reaches its condemning condition.
The “False Reset”: A Reservoir Drain Does Not Remove All of the Old Oil
The distinction between hydraulic-tank capacity and total hydraulic-system capacity is important.
For the R9800:
Hydraulic tank: approximately 5,800 L
Total hydraulic system: approximately 10,000 L
The difference is approximately:
4,200 litres.
That oil is distributed throughout the hydraulic system—in pumps, valves, manifolds, cylinders, motors, coolers, hoses, pipes and other circuit volumes.
Therefore, simply draining the hydraulic reservoir does not necessarily remove all of the degraded or contaminated oil.
Depending on machine position and the service procedure used, 42% of the total hydraulic oil system volume is residual inventory that remains trapped throughout the system.
Conceptually:
10,000-L total hydraulic system
− 5,800-L reservoir drain
≈ 4,200 L potentially remaining outside the reservoir
That represents approximately 42% of total nominal system capacity.
The precise quantity remaining during an actual service event will vary with cylinder positions, circuit configuration, draining procedure and other factors. Nevertheless, the basic engineering issue remains:
New oil is being mixed with residual used oil.
If that residual oil contains oxidation products, varnish precursors or other contamination, a reservoir drain does not provide a complete chemical reset.
It dilutes the contamination.
It does not necessarily eliminate it.
This is particularly important with varnish because deposit-forming material is not confined to the free oil in the reservoir. Degradation products and deposits may already exist on valve surfaces, hoses, manifolds, cooler surfaces and other components.
The machine can therefore return to service with thousands of litres of residual oil and contaminated wetted surfaces immediately mixing with the new charge.
The Economic Problem with Calendar-Based Oil Replacement
Consider an R9800 operating approximately 8,000 hours per year.
At a 1,000-hour hydraulic-oil replacement interval, the machine experiences approximately:
Eight hydraulic drain events per year.
Now compare that with a SmartOil G3 condition-based strategy in which the hydraulic fluid is continuously monitored, supported by Exception Sampling™ and laboratory analysis, with a planned full reservoir drain only twice per year, subject to the fluid remaining within its approved condition limits.
That reduces scheduled drain events from:
8 per year → 2 per year
or:
Six avoided hydraulic-oil drain events per year.
The economic value is not primarily the cost of the oil.
It is the productive time required to service the machine.
Pilbara Iron-Ore Economics
Fortescue (FMG:ASX) reported a FY26 hematite realised price of approximately US$90.70 per dry metric tonne, with FY26 underlying EBITDA of US$8.635 billion and an underlying EBITDA margin of approximately 51%.
Public planning documents indicate an approximately 55-unit large hydraulic-excavator fleet, principally in the 400- and 600-tonne classes, while Fortescue's electrification program is progressively replacing that fleet. Fortescue does not publicly disclose a reliable current count by R9400, R9600, R9800, CAT 6040 and CAT 6060, so the fleet mix below is modeled rather than claimed as an FMG asset roster.
For purposes of illustrating the economic exposure of an R9800 loading iron ore, assume:
Effective ore loading rate: 6,000 dmt/hour
Fortescue FY26 realised hematite price: US$90.70/dmt
Underlying EBITDA margin proxy: 51%
Scheduled R9800 utilisation: 8,000 operating hours/year
Base hydraulic service downtime: 8 hours/event
The 6,000-dmt/hour production rate is an illustrative economic-model assumption rather than a claim for a particular Fortescue excavator.
At these assumptions:
Gross ore-value throughput
6,000 dmt/h × US$90.70/dmt
= approximately
US$544,200 per productive hour.
Applying the 51% EBITDA-margin proxy:
Approximately US$277,500 of EBITDA-equivalent production value per hour.
These figures represent production value exposed, rather than a claim that every hour of excavator downtime produces an identical permanent accounting loss.
A mine may recover some lost tonnes through stockpiles, dispatch changes, additional shifts or other production assets.
But where the excavator is the production constraint, the economic leverage is substantial.
Per-machine economics
By scaling the existing R9800 base model by nominal face-shovel capacity, Liebherr lists approximately 22 m³ for the R9400, 37 m³ for the R9600 and 42 m³ for the R9800; Caterpillar lists approximately 22 m³ for the CAT 6040 and 34 m³ for the CAT 6060.
The common assumptions remain 8,000 operating hours/year, eight 1,000-hour drain events under the conventional model, two drains/year under condition-based management, and eight hours of downtime per event. That yields 48 recovered productive hours per machine per year.
The throughput figures are economic-model proxies, not OEM production guarantees. Actual tonnes/hour depend heavily on bucket fill, material density, fragmentation, swing angle, bench configuration, truck match and operator practices.
Fortescue fleet-level impact
Because the exact model mix is not public, a useful way to frame the answer is with a 55-machine sensitivity.
The ~US$1.03 billion / ~US$523 million middle case as the most appropriate number for the Part 3 business argument. A nominal 30 m³ fleet-average bucket is consistent with a fleet dominated by R9400/CAT 6040 and R9600/CAT 6060 class machines, with a smaller population of R9800-scale units.
That theoretical US$523 million EBITDA-equivalent exposure equals about 6.1% of Fortescue's entire FY26 Underlying EBITDA of US$8.635 billion. That comparison demonstrates why hydraulic-fluid management can become an enterprise-level issue rather than merely a lubrication-maintenance initiative.
What FMG would realistically capture
The full US$523 million is not necessarily the incremental EBITDA in a commercial proposal. Several effects reduce the amount ultimately monetized: some excavators move waste rather than saleable ore; planned drain work can sometimes be coordinated with other maintenance; backup excavators and stockpiles can cushion an outage; and some lost production can be recovered later.
A useful realization factor is therefore:
Therefore, SmartOil G3 Adaptive Dosing™ has a credible modeled EBITDA opportunity of approximately US$180–260 million per year across the excavator fleet, with US$523 million/year representing the theoretical EBITDA-equivalent production exposure protected.
That US$180–260 million range is equivalent to roughly 2.1–3.0% of Fortescue's FY26 company-wide EBITDA—before adding the direct maintenance savings from reduced hydraulic-oil purchases, filters, disposal, service labor, oil transport, maintenance equipment and safety exposure.
Fortescue is actively electrifying its excavator fleet. It reported 16 electric excavators already operating in 2026 and stated that around half the fleet is expected to be electric by the end of 2026. Electrification does not eliminate the hydraulic-fluid opportunity. The R9400 E and future electric hydraulic excavators still depend on large, high-performance hydraulic systems. Therefore, the SmartOil G3 / DECON / Adaptive Dosing™ value proposition can remain relevant through Fortescue's diesel-to-electric fleet transition.
Fleet Scale: The Potential EBITDA Impact Across Fortescue's Excavator Fleet
The economics become substantially more important when the analysis moves from a single R9800 to the entire Fortescue hydraulic-excavator fleet.
“Across an approximately 55-unit Fortescue hydraulic-excavator fleet, moving from fixed 1,000-hour reservoir drains to validated condition-based fluid management could protect approximately US$1.0 billion of annual production value and US$523 million of EBITDA-equivalent production exposure. Even at only 35–50% economic realization, the modeled annual EBITDA opportunity is approximately US$180–260 million.”
Fortescue's published decarbonisation planning identifies an approximate fleet quantity of 55 large hydraulic excavators in the 400- and 600-tonne classes. Fortescue's replacement strategy calls for this population to progressively transition to Liebherr R9400 E and R9600 E electric excavators.
The existing and transitional digging fleet includes a mix of large Liebherr and Caterpillar hydraulic excavators, including machines in the R9400, R9600, R9800, CAT 6040 and CAT 6060 classes. Fortescue does not publicly disclose a current model-by-model census for those machines. The fleet-level economic model therefore uses the 55-excavator population as the published fleet-scale reference and applies a blended productivity assumption rather than implying a precise current model mix.
For the base case, assume an average excavator productivity equivalent to approximately a 30 m³ bucket class, or about 4,286 dry metric tonnes per productive loading hour when scaled from the 6,000 dmt/h R9800 model.
Fortescue reported FY26 hematite realized pricing of US$90.70/dmt, record shipments of 201.3 Mt, Underlying EBITDA of US$8.635 billion, and an Underlying EBITDA margin of 51%.
At those economics, the modeled blended excavator supports approximately:
US$389,000 of gross iron-ore production value per productive hour
and approximately:
US$198,000 of EBITDA-equivalent production value per productive hour.
The significance of SmartOil G3™ Adaptive Dosing™ therefore increases dramatically when the potential recovery of productive hours is applied across the full excavator fleet.
55-Excavator Fleet Economic Model
The fleet model uses the same maintenance assumptions established for the R9800 analysis:
Scenario A — Fixed 1,000-hour drain
Eight hydraulic-oil drain events per excavator per year, assuming 8,000 operating hours annually and eight hours of downtime per drain event.
Scenario B — Condition-Based Fluid Management
Two planned hydraulic-oil drain events per excavator per year, supported by SmartOil G3™ multi-domain fluid sensing, Exception Sampling™, laboratory confirmation and Adaptive Dosing™.
The change reduces scheduled drain events from eight to two per machine and potentially returns:
48 productive hours per excavator per year.
Across 55 excavators:
2,640 productive excavator-hours may be returned annually.
Fortescue 55-Excavator Fleet Economics
The result is economically significant.
Under the modeled assumptions, shifting the 55-machine excavator fleet from eight scheduled hydraulic-oil drains per year to two condition-based drains potentially protects approximately:
US$1.03 BILLION
of annual gross iron-ore production-value exposure.
At Fortescue's FY26 51% Underlying EBITDA margin, this represents approximately:
US$523 MILLION
of annual EBITDA-equivalent production exposure.
That theoretical value is equivalent to approximately 6% of Fortescue's entire FY26 Underlying EBITDA of US$8.635 billion.
EBITDA Exposure Is Not the Same as Guaranteed EBITDA
The US$523 million figure should not be represented as a guaranteed incremental improvement in Fortescue's reported EBITDA.
It represents the EBITDA-equivalent economic value associated with the 2,640 modeled productive excavator-hours that are no longer consumed by six additional scheduled hydraulic-oil drain events per machine.
The amount Fortescue ultimately converts into additional EBITDA depends on mine-specific operating conditions.
Some excavators may be moving waste rather than saleable ore. Maintenance may sometimes be coordinated with other planned outages. Stockpiles may temporarily buffer lost digging capacity. Fleet dispatch may move trucks to another loading unit. Lost tonnes may sometimes be recovered later.
For this reason, a realization factor provides a more appropriate enterprise-level financial range.
A reasonable commercial planning range is therefore approximately:
US$180–260 MILLION PER YEAR
of potential realized EBITDA impact if Fortescue captures approximately 35–50% of the modeled productive-hour value.
That represents approximately 2–3% of Fortescue's FY26 company-wide Underlying EBITDA.
And this analysis considers only the production value associated with scheduled hydraulic-oil drain downtime.
It does not yet include the direct operating savings from reduced hydraulic-oil consumption, fewer filter replacements, reduced waste-oil disposal, lower maintenance labor, reduced service-equipment requirements or reduced maintenance-related personnel exposure.
Nor does it assign additional value to avoiding an unplanned varnish-related hydraulic failure.
The Fleet Multiplier Changes the Investment Decision
At the individual-machine level, SmartOil G3™ Adaptive Dosing™ can be evaluated as a hydraulic-fluid reliability technology.
At the 55-excavator fleet level, it becomes a mine-productivity and EBITDA technology.
The economic progression is:
Better fluid-condition visibility
→ fewer unnecessary hydraulic-oil drains
→ fewer maintenance interventions
→ more available excavator hours
→ more tonnes moved
→ greater utilization of the haulage fleet
→ greater production capacity
→ higher potential EBITDA.
This distinction is important because the cost of hydraulic oil is not the primary economic driver.
The productive hour is.
Across 55 large mining excavators, eliminating only six unnecessary eight-hour hydraulic-oil drain events per machine represents 330 avoided maintenance events and 2,640 productive hours potentially returned to Fortescue's mining system every year.
That is the scale at which condition-based hydraulic-fluid management should be evaluated.
SmartOil G3™ is not simply extending hydraulic-oil life.
It is protecting excavator availability—and converting fluid intelligence into potential enterprise EBITDA.
Recommended SmartOil G3™ Hardware Configuration for the R9800
Implementing Autonomous Fluid Intelligence™ on an R9800 requires more than a single oil-condition sensor or an automated sampling valve. The recommended SmartOil G3™ architecture combines multi-domain fluid sensing, autonomous physical sampling, Exception Sampling™, Adaptive Dosing™, machine operating data, power conditioning and edge computing into a single integrated machine-level system.
The architecture is designed around four primary functions:
1. Detect — continuously monitor hydraulic-fluid condition and machine operating state.
2. Confirm — automatically capture a representative physical oil sample when abnormal conditions warrant laboratory analysis.
3. Act — accurately deliver an approved quantity of Fluitec DECON™ when treatment is justified.
4. Verify — monitor the post-treatment response and determine whether the fluid is returning toward its approved performance envelope.
At the center of the architecture is the G3 Edge-AI Brain™, which receives sensor data, maintains the hydraulic-fluid and DECON mass balance, evaluates multi-domain trends, determines when an Exception Sample™ should be captured and controls the G3 DOSE™ module within validated dosing limits.
For an R9800, the recommended configuration is as follows.
The G3 Edge-AI Brain™ as the System Controller
The G3 Edge-AI Brain™ is the coordinating intelligence layer.
It does not simply collect sensor readings.
It combines independent fluid-condition measurements with machine operating context and evaluates how those variables change together over time.
For example, an isolated increase in dielectric constant may not justify intervention.
However:
**increasing dielectric deviation
-
increasing viscosity
-
worsening Tan Delta/EIS response
-
elevated hydraulic temperature
-
increasing filter differential pressure**
creates a much stronger indication that the hydraulic fluid is moving away from its established healthy baseline.
This sensor-fusion architecture is fundamental to SmartOil G3.
The Edge-AI Brain™ should therefore not command an Exception Sample™ or DECON dose solely because one sensor crosses one numerical threshold.
Instead, it evaluates:
-
absolute measurements;
-
temperature-normalized measurements;
-
rate of change;
-
persistence of the abnormal condition;
-
agreement among independent sensor domains;
-
machine load and operating state;
-
previous laboratory results;
-
oil-addition history;
-
previous DECON dosing;
-
time since the previous intervention; and
-
previous post-dose response.
The result is a progressively higher-confidence assessment of whether the observed condition represents normal variation, contamination, lubricant degradation, an emerging mechanical problem or developing varnish potential.
Autonomous Sampling and Exception Sampling™
The G3-HYD™ module provides the physical connection between continuous sensor intelligence and conventional laboratory oil analysis.
The preferred installation uses a representative circulating-oil bypass loop rather than taking stagnant oil from the bottom of the reservoir.
Oil is conditioned to an appropriate sampling pressure and flow before entering the G3-HYD™ manifold.
The module can operate in two modes.
Routine Autonomous Sampling
A physical sample can be collected automatically at an approved operating-hour or calendar interval to maintain continuity with the mine's existing oil-analysis program.
Exception Sampling™
When the Edge-AI Brain™ identifies a sufficiently abnormal multi-domain condition, the G3-HYD™ module captures a physical sample while the abnormal event is occurring.
The sample is automatically associated with:
-
machine identification;
-
operating hours;
-
date and time;
-
sensor measurements;
-
machine operating state;
-
anomaly type;
-
previous oil additions;
-
previous DECON additions; and
-
the logic that initiated the sample.
The sample can then be sent to the mine's approved laboratory for confirmatory testing such as:
MPC + RULER + FTIR + TAN + viscosity + Karl Fischer water + ISO 4406 + PQ + ICP elemental analysis.
This preserves the role of the laboratory while substantially improving when the sample is collected.
G3 DOSE™: The Controlled Actuation Layer
The G3 DOSE™ module is the physical actuation component of Adaptive Dosing™.
For the R9800 varnish application, the module draws Fluitec DECON AW from a dedicated on-machine treatment reservoir and delivers it into an approved hydraulic-oil mixing point.
The target is not simply to inject a predetermined volume periodically.
Instead, the Edge-AI Brain™ maintains an estimated DECON mass balance based on:
**total hydraulic-system inventory
-
hydraulic-oil additions
− hydraulic-oil losses
− drain events
-
previous DECON additions
− DECON removed with lost or drained oil.**
For the approximately 10,000-L R9800 hydraulic system, a Fluitec-approved concentration in the 3–5 vol% treatment range corresponds theoretically to approximately 300–500 L of DECON distributed through the complete circulating oil inventory.
Once the initial treatment concentration has been established, G3 DOSE™ is primarily intended to make incremental corrective additions required to compensate for dilution, oil loss and other changes in the fluid inventory.
An independent dosing-flow or volume-verification channel is recommended so the system can establish:
Commanded dose versus actual delivered dose.
This is important for creating a defensible treatment record and preventing accumulated dosing error.
The G3 Power Module: Machine-Grade Power Management
The G3 Power Module provides the electrical foundation for the complete installation.
Large mining machines present a demanding electrical environment that may include voltage transients, starting events, electrical noise, vibration and varying machine states.
The G3 Power Module should therefore provide the protected interface between the R9800 electrical system and the SmartOil G3 equipment.
Its recommended functions include:
-
protected machine DC input;
-
over-current protection;
-
reverse-polarity protection;
-
transient and surge suppression;
-
regulated power conversion where required;
-
protected power distribution to sensors and modules;
-
controlled outputs for solenoid valves and dosing hardware;
-
ignition/key-state input;
-
power-state monitoring;
-
fault reporting to the Edge-AI Brain™; and
-
controlled shutdown or restart behavior.
This allows the Edge-AI Brain™, sensing modules, sampling valves and dosing equipment to function as one integrated machine system rather than as independent electrical accessories.
One Architecture — Four Closed Loops
The recommended SmartOil G3 R9800 configuration can be viewed as four interconnected loops.
Fluid Intelligence Loop
G3-SEN1™ + G3-SEN2™ + particle count + water/EIS + temperature + ΔP
→ G3 Edge-AI Brain™
→ anomaly probability and fluid-health trend.
Physical Verification Loop
Anomaly detected
→ Exception Sampling™
→ G3-HYD™
→ physical oil sample
→ laboratory confirmation.
Adaptive Treatment Loop
Confirmed varnish/degradation condition
→ Edge-AI mass-balance calculation
→ G3 DOSE™
→ controlled DECON AW addition
→ circulation
→ sensor and laboratory verification.
Economic Protection Loop
Improved fluid stability
→ fewer unnecessary oil drains
→ fewer maintenance interventions
→ greater R9800 availability
→ more productive loading hours
→ more tonnes moved
→ protected EBITDA.
Together, these components transform the hydraulic oil from a periodically sampled maintenance consumable into a continuously monitored and actively managed production variable.
That is the practical hardware foundation of SmartOil G3™ Autonomous Fluid Intelligence™:
Sense continuously.
Sample when the evidence warrants it.
Dose only when justified.
Verify the response.
Protect the productive hour.
Scenario A vs. Scenario B
Annualized R9800 Hydraulic-Oil Strategy
Base-case economic result
Moving from eight scheduled hydraulic drain events to two condition-based drain events potentially returns:
48 productive hours per year.
At the illustrative Pilbara production economics used above, those 48 hours represent approximately:
US$26.1 million of gross ore-production value
and approximately
US$13.3 million of EBITDA-equivalent production exposure.
This excludes the additional direct savings associated with:
-
approximately 34,800 fewer litres of new hydraulic oil consumed annually;
-
six fewer hydraulic-filter replacement events;
-
reduced oil disposal;
-
reduced service labour;
-
reduced mobile-maintenance support;
-
reduced maintenance-related safety exposure; and
-
reduced risk associated with opening and servicing a large hydraulic system.
The economics therefore have two layers:
Direct maintenance savings
plus
productive machine availability.
For an R9800, the second category can overwhelm the first.
Downtime Sensitivity
The actual time required to completely drain, service, refill, circulate, inspect and return an R9800 to production will vary by mine.
The economic model should therefore be considered across several service-duration assumptions.
Even at only four hours of downtime per drain event, the economic value associated with six avoided interventions remains material.
This is why the business case should not be framed around:
“How much does the hydraulic oil cost?”
The more important question is:
“How much excavator production is consumed by changing it?”
Condition-Based Oil Life Versus Scheduled Disposal
A fixed 1,000-hour drain interval assumes that hydraulic-oil age is an adequate proxy for hydraulic-oil condition.
It often is not.
Two identical R9800 excavators may accumulate 1,000 hours under very different conditions.
One may operate with stable hydraulic temperatures, effective cooling, minimal oil makeup and low contamination.
The other may experience recurring thermal excursions, heavy loading, frequent hose replacements or abnormal contamination.
Under a purely hour-based maintenance model, both machines reach the same maintenance decision:
Drain the oil.
Condition-based monitoring asks a different question:
Does this particular oil actually need to be replaced?
SmartOil G3 continuously monitors the fluid and combines sensor measurements with operating context.
Laboratory analysis remains an important verification tool.
Liebherr itself identifies used-oil analysis as a means of extending hydraulic-oil utilization while maintaining machine reliability and optimizing operating cost.
That creates an important distinction.
The objective of SmartOil G3 is not simply to extend oil life.
The objective is to safely extend oil life when the evidence demonstrates that the oil remains suitable for service.
Exception Sampling™: Sample the Event, Not Just the Calendar
Traditional scheduled oil analysis may collect a sample every 250, 500 or 1,000 operating hours.
That produces a valuable trend.
But it cannot observe every abnormal thermal event, cooler problem, heavy-load episode, oil addition or contamination event that occurs between samples.
The condition that matters most may occur hundreds of hours before the next scheduled sample.
Exception Sampling™ changes that relationship.
When the G3 Edge-AI Brain™ identifies an abnormal fluid or machine condition, a physical sample can be captured while the abnormal condition is occurring.
This produces something routine sampling cannot consistently provide:
A laboratory sample tied directly to the event.
The laboratory can then evaluate MPC, antioxidant condition and other appropriate fluid-health parameters.
The result is a much stronger relationship between:
machine event → sensor response → physical sample → laboratory evidence → maintenance decision.
Adaptive Dosing™: Treat the Condition Without Masking the Cause
A developing varnish condition does not automatically mean that treatment chemistry should be added.
An increase in treatment demand may itself be diagnostic.
A recurring increase in varnish potential could indicate sustained elevated temperature caused by reduced cooler performance.
Repeated hydraulic-oil additions can change both contaminant and treatment concentration.
A cooling-fan problem may accelerate oxidation even while conventional viscosity remains within limits.
SmartOil G3 must therefore distinguish between:
a fluid condition that can appropriately be treated
and
a mechanical condition requiring maintenance intervention.
Adaptive Dosing™ treats the fluid while Autonomous Fluid Intelligence™ helps determine why treatment demand changed.
That distinction prevents chemistry from masking a mechanical problem.
The Economics of Prevention Are Asymmetric
The direct cost of managing 5,800 litres of reservoir oil is significant in conventional maintenance terms.
But it remains small compared with the production value supported by the hydraulic system.
The R9800’s approximately 10,000-L hydraulic circuit controls an ultra-class excavator whose productive hour can represent hundreds of thousands of dollars of ore throughput.
This creates an asymmetric reliability proposition.
SmartOil G3 does not need to double hydraulic-oil life on every machine to create value.
It does not need to eliminate every varnish event.
It needs to help the mine make better decisions about:
when the oil actually requires intervention;
when it remains fit for continued service;
when an abnormal condition requires a physical sample;
when treatment is appropriate;
and
when the problem is actually mechanical rather than chemical.
Avoiding even a small number of unnecessary major maintenance interventions can produce an economic return many times greater than the direct cost of the fluid-management program.
From Pilot to Bounded Autonomy
A reliable implementation should begin with a bounded engineering pilot.
Select an R9800 with meaningful production exposure and sufficient operating and oil-analysis history.
Preference should be given to a machine showing one or more of the following:
-
elevated or unstable hydraulic temperature;
-
recurring filter issues;
-
accelerated antioxidant depletion;
-
elevated MPC or varnish potential;
-
repeated hydraulic-oil replacement;
-
recurring valve problems; or
-
unexplained changes in hydraulic performance.
Document:
-
oil formulation;
-
approximately 5,800-L reservoir volume;
-
approximately 10,000-L total system inventory;
-
filtration configuration;
-
oil-addition history;
-
cooler performance;
-
component history;
-
machine operating profile; and
-
historical laboratory results.
Establish baseline laboratory samples before enabling dosing.
The program should then progress through four controlled stages.
Monitoring Mode
SmartOil G3 characterizes normal relationships among fluid condition, hydraulic temperature, load, machine hours and oil additions.
Advisory Mode
The system identifies developing abnormal conditions, initiates Exception Sampling™ when appropriate and recommends intervention for engineering approval.
Supervised Dosing Mode
Approved logic commands a bounded treatment dose while personnel review the event and subsequent fluid response.
Autonomous Mode
The system acts within validated limits, documents every intervention and escalates conditions outside the approved model.
Autonomy should be earned through evidence.
Measure the Business Result, Not Just the Oil Result
A successful mining pilot should measure three categories of performance.
Fluid Metrics
-
MPC trend
-
antioxidant reserve
-
viscosity
-
dielectric trend
-
acid number
-
particle behavior
-
water
-
filter differential pressure
-
oil makeup
-
treatment concentration
Machine Metrics
-
hydraulic temperature
-
valve response
-
hydraulic alarms
-
filter life
-
cycle-time consistency
-
pump or valve interventions
-
unscheduled hydraulic maintenance
Business Metrics
-
excavator availability
-
scheduled drain downtime
-
unscheduled hydraulic downtime
-
litres of hydraulic oil consumed
-
filter-service events
-
maintenance labour
-
tonnes moved
-
truck queueing
-
cost per tonne
-
production hours preserved
-
EBITDA-equivalent production exposure protected
The objective is to establish a defensible relationship:
Fluid condition → intervention → hydraulic response → machine response → production result.
Creating a Defensible Chain of Evidence
The strongest proof of value will come from combining a fluid mass balance with machine-production data.
How much oil was added?
How much was drained?
How much treatment was delivered?
What volume of residual oil remained in the machine?
What was the varnish-potential trend?
How did antioxidant reserve respond?
How did hydraulic temperature change?
Did filter differential pressure improve?
Did valve interventions decline?
Were fewer premature drain events required?
How many maintenance hours were eliminated?
How many productive hours were returned to the machine?
And ultimately:
How many tonnes remained in the mine plan because the excavator remained available?
This is where SmartOil G3 moves beyond conventional oil analysis.
The larger shift is from periodic oil maintenance to Autonomous Fluid Intelligence™.
And in mobile mining, the final step is from Autonomous Fluid Intelligence™ to:
Production protection.
The Machine Is the Business Case
Varnish will never be eliminated as a chemical possibility.
Hydraulic oils will continue to oxidize.
Machines will generate heat.
Hoses will fail.
Oil will be added.
Filters will load.
Operating conditions will change.
The objective is not to promise permanently perfect hydraulic oil.
The objective is to keep the fluid inside its approved performance envelope and intervene before deterioration becomes a machine-performance problem.
For an R9800 operating in the Pilbara, this distinction has major economic implications.
The machine contains approximately 10,000 litres of hydraulic fluid, but only about 5,800 litres are contained in the hydraulic tank.
Simply draining the tank does not eliminate all of the old oil or the varnish precursors distributed throughout the hydraulic circuit.
Replacing the oil more frequently is therefore not necessarily the same as managing the oil more intelligently.
At contemporary Fortescue iron-ore economics, every productive R9800 hour may represent hundreds of thousands of dollars of ore throughput.
That is the economic argument for SmartOil G3 Adaptive Dosing™ and Autonomous Fluid Intelligence™:
Do not change the oil simply because the calendar says it is old.
Determine whether it is healthy.
Capture the exception when it occurs.
Treat only when justified.
Drain when condition requires it.
And protect the productive hour.
Conclusion
Varnish control on ultra-class mining equipment is not fundamentally an oil-management problem.
It is an availability and production problem.
On a machine such as the Liebherr R9800, hydraulic-fluid condition directly influences valve response, thermal stability, filtration performance, component reliability, loading-cycle consistency and ultimately the number of productive hours the machine can deliver. In the Pilbara iron-ore environment, those productive hours carry significant economic value.
The traditional response to hydraulic-oil degradation risk has often been to manage uncertainty with fixed maintenance intervals: drain the reservoir, replace the filters, refill the machine and return it to service.
That approach is simple, but it is also blunt.
A 1,000-hour interval cannot determine whether the oil is actually degraded, whether varnish potential is increasing, whether contamination is present, or whether the underlying problem is mechanical rather than chemical. Nor does draining the reservoir completely reset the hydraulic system. A substantial quantity of used oil remains distributed throughout valves, pumps, cylinders, coolers, motors, hoses and piping, where degradation products and deposits may persist.
SmartOil G3™ introduces a different operating model.
The G3-SEN™ multi-domain sensor architecture continuously observes the hydraulic fluid. The G3 Edge-AI Brain™interprets those measurements in the context of machine state, operating history, temperature, oil additions, contamination and prior treatment. Sensor fusion increases confidence by correlating multiple independent signals rather than relying on a single threshold crossing.
When abnormal behavior is detected, Exception Sampling™ captures the physical condition while the event is occurring. Laboratory analysis then provides the confirmation layer required for higher-order measurements such as MPC varnish potential, antioxidant reserve, oxidation, TAN, water, particle contamination and wear debris.
When varnish mitigation is justified, Adaptive Dosing™ closes the loop.
Using the G3 DOSE™ module and approved Fluitec DECON™ chemistry, the system can calculate and deliver a controlled treatment quantity while maintaining a running fluid and treatment mass balance. Rather than treating DECON as a periodic manual additive, SmartOil G3 can manage it as a controlled condition-based variable—accounting for oil makeup, leakage, drains, refills and prior dosing while maintaining the validated treatment concentration.
The result is a fundamentally different control strategy:
Measure → Interpret → Correlate → Exception Sample™ → Confirm → Calculate → Dose → Circulate → Verify → Repeat only when necessary.
This architecture does not eliminate the laboratory, maintenance personnel or engineering judgment.
It makes each of them more effective.
The laboratory receives a sample taken at the moment it matters.
Maintenance personnel intervene when condition justifies the work.
Reliability engineers receive a continuous record of fluid condition, treatment history and machine response.
And the mine gains the evidence required to determine whether the hydraulic oil should continue in service, be filtered, treated, repaired around, or drained.
The economic consequence is significant.
At the individual-machine level, condition-based hydraulic-fluid management can reduce unnecessary drain events, lower oil consumption, reduce filter use, decrease maintenance labor and reduce personnel exposure to large-volume oil and filter-change activities.
At the fleet level, the value multiplies.
For Fortescue's approximately 55 large hydraulic excavators, the modeled reduction from eight scheduled drain events per year to two condition-based events represents 2,640 productive excavator-hours potentially returned to the mining system annually. Under the Pilbara economic assumptions developed in this article, that corresponds to approximately US$1.03 billion of gross production-value exposure and about US$523 million of EBITDA-equivalent exposure.
Those figures are not a guarantee of incremental EBITDA. Actual value realized depends on mine scheduling, asset utilization, ore versus waste movement, maintenance coordination and the ability to convert recovered machine availability into additional tonnes.
But the scale of the opportunity is clear.
Even partial realization of the recovered productive time can create an enterprise-level economic impact far greater than the direct cost of hydraulic oil itself.
That is the larger message of this three-part varnish series.
Varnish begins as a chemical degradation mechanism.
It becomes a fluid-condition problem.
If unmanaged, it can become a hydraulic-performance problem.
And on a high-value mining asset, it can ultimately become a production and EBITDA problem.
SmartOil G3™ is designed to intervene earlier in that progression.
Sense the condition.
Capture the evidence.
Treat only when justified.
Verify the response.
Extend oil life when the data supports it.
And protect the productive hour.
That is the transition from periodic oil maintenance to Autonomous Fluid Intelligence™—and from fluid management to production protection.
About LogiLube
LogiLube, LLC is a Denver, Colorado-based technology company focused on advancing real-time fluid condition monitoring and predictive maintenance through its patented SmartOil® platform. The company’s SmartOil G3 Autonomous Fluid Intelligence™ layer combines 3rd party in situ sensors, Edge-AI processing, and Local Language Models (LoLM) to deliver continuous monitoring, anomaly detection, and remaining useful life (RUL) predictions for industrial assets. Serving industries such as data centers, mining, and energy, LogiLube enables operators, OEMs, and asset owners to reduce unplanned downtime, optimize maintenance strategies, and unlock the value of high-fidelity operational data across distributed fleets.
SmartOil G3™ technology is protected by U.S. Patent No. 10,466,152; 11,761,946; 12,681,003; International Patents, and other U.S. and International Patents Pending
Copyright ©2026 LogiLube, LLC All Rights Reserved