Enhancing project delivery through operational experience
How integrating operational insight early in EPCM project design reduces commissioning risk, CAPEX overruns and ramp-up duration in mining and mineral processing
ABOUT THIS PAPER
What this paper covers - and why it matters for your next EPCM project
Engineering, procurement and construction management (EPCM) remains the dominant delivery model for greenfield mining projects. It provides structure, accountability and access to specialist capability. But the rigid phase segmentation – engineering, construction, commissioning, operations creates a persistent knowledge gap.
When operational stakeholders are introduced late, problems with accessibility, maintainability, instrumentation placement, control logic and metallurgical assumptions are only discovered after construction is complete. Workarounds mount. Scope grows. Ramp-up drags. CAPEX blows out.
“Design deficiencies can be repeated and addressed late during commissioning, increasing ramp-up duration and elevating CAPEX risk which can undermine investor confidence.”
This paper argues for a more iterative, experience-driven approach – one that integrates operational and commissioning expertise from the outset and institutionalises feedback from completed projects into subsequent designs.
WHO SHOULD READ THIS
This paper and presentation are written by practitioners, for practitioners. If you’re involved in the design, delivery or operation of mineral processing infrastructure, the arguments here are relevant to your work.
will recognise the CAPEX and schedule dynamics described, and find a framework for challenging the status quo in how operational input is structured into EPCM contracts.
will find specific, actionable examples of where operational knowledge changes design decisions, from sump routing and probe placement to control architecture and alarm philosophy.
working in project or operational roles will find the testwork critique and geometallurgical recommendations directly applicable to how variability is handled in current or upcoming projects.
who’ve lived the consequences of late-stage discovery will find their experience validated, and a structured case for why earlier involvement produces better outcomes for everyone.
will find the digital design section relevant, particularly the argument that instrumentation and control logic decisions should be driven by process understanding rather than technology preference.
If you’ve ever walked into a newly constructed plant and immediately spotted something that should have been obvious in design – this paper is for you.
KEY THEMES
The six EPCM failure modes this paper addresses
The paper is structured around five interconnected themes, each addressing a distinct failure mode in conventional EPCM project delivery.
THEME 1
Why EPCM's linear phase structure creates commissioning risks
The linear EPCM delivery model creates rigid segmentation. Design influence is highest – and cheapest to act on – in the pre-feasibility and FEED phases, yet operational stakeholders are typically introduced only at the 20–25% engineering complete stage.
By this point, the cost of change has risen significantly. The paper maps the compounding cost of late discovery, drawing on Boyd’s (1976) cost-of-change curve.
THEME 2
How operational experience in mining plant design prevents costly rework
Operational experience – from plant operators, metallurgists, maintainers and control technicians – asks questions that don’t appear on a datasheet. Examples from the paper include:
- Sump pump routing configured to prevent bogging due to cyclical pump operation
- pH probe placement balancing maintenance access with adequate reagent conditioning time
- Weigh to meter location for control responsiveness and maintainability
- Platform positioning to give supervisors visual coverage of upstream and downstream processes
- Fieldbus and I/O architecture choices that affect long-term diagnostics and expandability
THEME 3
Why bulk composite test work fails real ore variability - and what to do instead
Orebodies are rarely homogenous. Designs based on idealised composite samples risk operational failures when actual ore variability is encountered in production. The paper advocates:
- Variability testing within geological domains, not just domain-average composites
- Geometallurgical domain ranking to prioritise test work investment
- Life-of-Mine schedule integration to expose periods where the plant design may underperform
- Blending sensitivity analysis to identify antagonistic ore type combinations
- Designing for ranges of conditions rather than single-point assumptions
THEME 4
Control system design in mineral processing: why metallurgical insight must come firsT
Control systems are frequently designed with technology, purchasing and security as the primary lens – rather than process understanding. The paper outlines how metallurgical insight should drive instrumentation selection, alarm philosophy and cascade control logic. It references the five-level Automation Pyramid (field instrumentation through to enterprise ERP) and argues that unless the pyramid is built on a solid, process-informed base, it will fail in practice.
The paper also addresses the Ridgeway ‘what gets measured gets managed’ principle – distinguishing between proxy measurements and the fundamental process variables that actually drive plant performance.
THEME 5
How to build operational feedback loops across EPCM project generations
Lessons learned from commissioning rarely find their way back into design. The paper proposes structured feedback pathways – from post-commissioning reviews to the integration of AI and LLMs trained on prior project libraries – to institutionalise continuous improvement across EPCM project generations.
THEME 6
Cross-functional collaboration in mining projects: the business case for change
The mining sector has historically siloed expertise – engineers engineer, operators operate, metallurgists analyse. The paper argues that cross-functional collaboration and earlier design reviews deliver returns that far outweigh the alternative: a project delivered on time but requiring a 40% capital blowout to fix post-construction problems.
FREE DOWNLOADS
Download the MetPlant 2026 paper and presentation - no registration required
Both artefacts from Drew’s MetPlant 2026 presentation are available for immediate download – no registration required.
TECHNICAL PAPER - PDF
Full peer-reviewed paper covering EPCM limitations, the value of operational expertise, metallurgical testing realities, control system philosophy and a framework for closing the operational feedback loop. Includes figures, tables and references.
CONFERENCE PRESENTATION - PDF
The presentation delivered by Drew Clements at MetPlant 2026. Covers catalyst for change, the core EPCM challenge, early operational involvement, digital design tools, real-world examples, empirical feedback frameworks and a reflection section for owners, designers and suppliers.
ABOUT THE AUTHORS
Mipac's Process Engineering and Metallurgy Team
This paper draws on decades of combined experience in mineral processing operations, metallurgical engineering and industrial automation.
Drew Clements
Senior Metallurgist Optimisation Team Lead
MAusIMM. Drew brings a rare breadth across metallurgy, operations, maintenance, management, automation and consulting. Currently leading Mipac’s optimisation practice.
Luke Evans
Senior Process Engineer, Optimisation Team
MAusIMM CP (Met). A Chartered Metallurgist with deep experience in plant design, commissioning support and process optimisation across commodity groups.
Dinoe Bertoli
Previously Senior Process Engineer, Optimisation Team
FAusIMM. A Fellow of the AusIMM with extensive experience in process engineering, bridging design intent and operational reality.
Ready to close the loop on your next project?
Whether you’re in feasibility, FEED or heading into detailed design, Mipac’s process engineers, metallurgists and control systems specialists can help you build operational intelligence into your project from day one.
Frequently Asked Questions
The linear EPCM delivery model creates rigid segmentation between engineering, construction, commissioning and operations phases. When operational stakeholders are involved late (typically at the 20–25% engineering complete stage) issues with accessibility, maintainability, instrumentation placement, control logic and metallurgical assumptions may only surface after construction is complete.
These late discoveries require expensive workarounds, scope growth and budget and schedule pressure that can undermine investor confidence and delay production ramp-up.
The optimal time for operational input is during the project feasibility study and Front-End Engineering Design (FEED) phase, when operator insights can be assessed via cost-benefit analysis without affecting project schedule. This is when design influence is highest and the cost of making changes is lowest.
A secondary intervention point is during project reviews prior to detailed design completion, to catch fatal flaws before they are locked into construction drawings.
By front-loading operational expertise into the design phase, projects move further along the experience curve before commissioning begins. The paper uses an Experience–Confidence framework to show that this means commissioning occurs in the ‘Slope of Enlightenment’ rather than the ‘Valley of Despair’ - resulting in faster ramp-up, fewer costly surprises and lower CAPEX risk.
Metallurgical testwork is the foundation of process design, driving equipment sizing, recovery targets, reagent selection and throughput predictions. Designs based on bulk composite samples or idealised ore assumptions risk operational failure once the full variability of the orebody is encountered. The paper advocates a geometallurgical approach: domain-based variability testing, Life-of-Mine schedule integration, blending sensitivity analysis and designing for ranges of conditions rather than single-point assumptions.
Control systems should be designed using metallurgical insight to build accurate processing models. This requires metallurgists and commissioning engineers to be involved in instrumentation selection, control philosophy development and alarm threshold setting. Key considerations include: feed-forward signalling from upstream processes, cascade control where warranted, removal of unnecessary process variability at source, and ensuring that what is measured is what actually matters to plant performance - not simply a proxy variable.
The paper identifies four compounding failure modes:
(1) a fragmented project lifecycle where phases have separate teams with different success metrics;
(2) misaligned incentives - engineers optimise for schedule and budget, operators prioritise safety and reliability;
(3) loss of tacit knowledge through high staff turnover and informal documentation; and
(4) digital tools that exist but lack the governance and cultural support to convert data into institutional learning.
Mipac integrates operational experience and subject matter expertise from feasibility study through FEED and into detailed design. This includes early engagement of experienced metallurgists, control system engineers and commissioning personnel.
Mipac also advocates for iterative digital process models updated with feedback from completed projects, and sees the industrialisation of operational expertise through AI and LLM tools as the next step in accelerating this feedback loop across project generations.
Want to read more on this topic?
This whitepaper will open your thinking on the control system design in a greenfield operation. Check out these articles that explore this topic in more detail.
Operational efficiency in mineral processing
Operational efficiency in mineral processing is increasingly defined by stability, control quality and decision confidence, not new equipment. As ore…
The hidden economics of control system design in mineral processing
Control system design is one of the most overlooked economic levers in mineral processing. When done well, it stabilises circuits,…
Designing tomorrow’s processing plants
Greenfield mining projects shape operational performance for decades. The most successful plants are not defined by equipment alone, but by…