The $2.2 Million Mistake You Don't Know You're Making with Your Epiroc Drills

The Day I Thought I Had It Figured Out

When I first started managing spare parts procurement for our Epiroc rig fleet—this was in 2021—I assumed the main performance killer was downtime. If a drill wasn't running, we weren't making money. Simple, right?

So I focused everything on part availability. Pushed our Epiroc dealer for faster shipping. Stockpiled consumables. Kept a surplus of rock drill components. My logic: machines running = productivity. Seemed bulletproof.

I was wrong. Spectacularly wrong.

By Q4 2022, we had a fleet utilization rate of 89%. Impressive on paper. Yet our cost per ton blasted was climbing month over month. That's the moment I realized: uptime isn't the whole equation. I'd been optimizing for the wrong metric.

Surface Problem: The Fragmentation Drift

The issue looked like a blasting problem. Operators were complaining about uneven rock fragmentation. More oversize boulders needing secondary breakage. Slower loading cycles. Higher explosives costs. We blamed the rock—it's a variable site, right?

We spent two months tweaking blast patterns and powder factors. Adjusted stemming. Changed initiation sequences. Results: marginal. Average fragmentation size actually increased—from 350mm to 420mm over that period. That's a consequential shift when your crusher feed expects a consistent profile.

Here's the kicker: the drill pattern itself was drifting. Not catastrophically. Just enough to shift burden and spacing by 5-8% on average. But over a 200-hole blast, that accumulates into a major fragmentation inconsistency.

Deep Cause: The Hidden Automation Creep

The root cause wasn't the geology. It wasn't the explosives. It was the subtle degradation of our Epiroc rigs' automated drilling parameters.

This is the part that took me months to uncover, because everyone—including me—assumes automation is set-and-forget. Your Epiroc rig has deep automation. Mobilaris integration. It adjusts feed, rotation, percussion in real time. But those adjustments are based on algorithms that can drift.

Here's what I eventually found:

• Consumable wear (especially on DTH hammers) was altering the relationship between feed pressure and actual penetration rate. The automation compensated, but the compensation drift wasn't being flagged.
• Software baseline parameters had been reset during a firmware update (circa early 2022). The new defaults were slightly different—more aggressive on penetration, less aggressive on hole straightness—and nobody documented the change.
• Operator overrides. Some crews, trying to improve cycle times, were manually overriding the automated depth settings. Their reasoning: 'holes aren't deep enough.' The result? Off-pattern positioning.

I'm not a drill automation engineer. But after digging through 18 months of Mobilaris data—(note to self: actually do this more often)—I noticed the pattern. The standard deviation in hole positioning had increased by 1.3% quarter over quarter. Imperceptible shift. Devastating cumulative effect.

The Real Cost of Ignoring Parameter Drift

Let me put some real numbers on this, because 'fragmentation inconsistency' sounds academic. It's not.

In 2022, we blasted approximately 1.2 million tons of material. Due to the drift, our average fragmentation size increased by roughly 70mm. That forced us into:

• An extra 450 hours of secondary breakage (rock breakers) across the year
• Increased crusher downtime due to blockages—about 120 hours that never appeared on the drill downtime report (because the drill was running fine, the crusher was the bottleneck)
• Higher explosives consumption: we were using roughly $0.12/ton more ANFO than benchmark. On 1.2 million tons, that's $144,000 wasted annually.

Adding it up: the cumulative cost of that automation drift was approximately $680,000 in 2022 alone. Over three years? Close to $2.2 million. All because we were looking at uptime and ignoring fragmentation consistency. That error affected every downstream process. $680,000—straight to waste. When I finally calculated that number, I felt sick.

"The wrong metric on a fleet of Epiroc rigs = $680K wasted + 3 weeks of credibility repair with site management."

What We Actually Changed (Short, Because You Already Get It)

Once the problem was clear, the fix was surprisingly straightforward. But it required a mental shift in how we measure fleet performance.

1. We stopped measuring 'uptime' as a standalone metric.
My experience is based on about 40 orders and 4 years managing our Epiroc fleet. I can't speak to how this applies to every operation—especially smaller sites with different automation levels—but for us, the switch was to measure blast quality consistency as a leading indicator. If fragmentation starts drifting, you've got a week, maybe two, to intervene before cost per ton spikes.

2. We implemented a monthly parameter baseline check.
—or rather, we actually followed the Epiroc service manual. It recommends recalibrating automation feed algorithms every 500 operating hours or after a major firmware update. We'd been doing it... 'approximately.' Now it's a hard milestone logged in the CMMS. (Should mention: the Epiroc service team can run this check remotely via Mobilaris—we weren't using that feature.)

3. We enforced the 'no manual override' rule.
Learned never to assume operators won't override automation after I found it was happening on 3 of 5 shifts. The solution wasn't discipline—it was retraining. We ran a half-day workshop on how automation drift detection works. Once crews understood the fragmentation impact, the overrides stopped. Mostly.

The Real KPI You Should Track

If you're managing an Epiroc fleet, here's my unsolicited advice: look at fragmentation uniformity index (FUI) week over week. Not just in absolute terms, but in trend. If FUI shifts by more than 5% in a month, something in your drilling parameter chain has drifted—consumable wear, software settings, or operator input. Don't blame the rock. It's almost never the rock.

I should add: I wrote this because I wasted three years optimizing for the wrong thing. I assumed 'machine running = good.' Turned out I was measuring machine time, not productivity. Our Epiroc drills were never the problem. Our metrics were.