Why Gold Miners Look Greener but Burn More Energy Per Ounce

#

Two numbers from the same report are telling opposite stories. Scope 2 emissions across the world’s major gold producers have now fallen for four consecutive years, the kind of trend line that looks like decarbonisation working. At the same time, energy intensity has climbed to the highest point ever recorded, sitting 47% above 2016 levels. How can a sector be using cleaner power and getting less efficient in the same breath?

That contradiction sits at the centre of the Metals Focus Gold ESG Focus 2026 report, which tracked 18 major producers across their 2025 performance. The timing matters. Institutional capital is increasingly screening gold miners on per-ounce intensity metrics rather than headline emissions totals, which means the answer to that question now shapes how these companies are evaluated and priced.

What the 2026 data actually shows: two metrics pointing in opposite directions

Start with the emissions picture, because that is the story most headlines lead with. Combined Scope 1 and Scope 2 emissions across the 18 producers held roughly flat at 29.9 Mt CO2e in 2025. Scope 2 emissions, the ones tied to purchased electricity, fell for the fourth year running as renewable projects and power purchase agreements reduced grid dependence. Scope 3 dropped 9%.

Read only those figures and the sector looks like it is decarbonising steadily.

Now change the lens to intensity, and the picture inverts. Average emissions intensity rose 7% year-on-year to 0.9 tCO2e per gold-equivalent ounce, the third consecutive annual increase. Energy intensity climbed 12% to 10.2 GJ per ounce, the highest figure in the entire dataset. Scope 1 emissions, tied directly to fuel burned on site, rose in 2025 after two straight years of decline.

Energy intensity of 10.2 GJ per ounce now sits 47% above 2016 levels. The sector is burning far more energy for every ounce it produces than it did a decade ago.

The single-year move is what makes this sharp. Energy intensity ran at roughly 9.3 GJ per ounce in 2024, so the jump to 10.2 GJ per ounce represents a steep deterioration in just twelve months. Total energy consumed reached 334 PJ, up 4.8%, and direct energy from diesel and fuel oil rose 17% to account for 71% of the total.

The gap between these two lenses is the whole point. An ESG assessment built on absolute emissions will show this sector improving. One built on intensity will show it going backwards. Knowing which lens is being applied to a producer is the first step to reading its sustainability claims accurately.

Institutional capital increasingly applies ESG capital screening that weighs intensity metrics alongside absolute totals, creating a growing divide between producers whose sustainability narratives survive close scrutiny and those whose reported progress unravels once production context is accounted for.

Why efficiency is falling even as green energy adoption rises

Why would a mine adding solar panels and signing renewable power deals become less efficient? The instinct is to call it a contradiction. It is not. It is the predictable outcome of forces that renewable energy simply cannot touch.

The first is ore-grade decline. As the richest ore is mined out, what remains holds less gold per tonne of rock, so miners must crush, grind, and process far more material to produce the same ounce. That processing burns energy regardless of where the electricity comes from. Metals Focus attributed much of the 3% intensity rise in 2024 directly to this dynamic.

The second is the fixed-load problem. A mine’s ventilation, dewatering, and hoisting systems draw a relatively fixed baseline of energy whether output is high or low. When gold production falls, that fixed energy gets spread across fewer ounces, so energy per ounce rises mechanically, without anything changing on the ground.

The fixed-load problem is where energy efficiency tools that target ventilation and dewatering systems have their most direct application, since reducing baseline power draw at depth lowers the denominator effect on per-ounce intensity even when production is flat.

Lower gold output is a primary driver of rising average emissions intensity across consecutive years, according to Sarah Tomlinson, mine supply director at Metals Focus.

The four structural mechanisms behind the paradox are worth isolating:

The portfolio point carries a specific risk that analysts call ESG arbitrage. A producer can sell off its dirtiest mines, watch its consolidated emissions total drop, and present that as progress, even though every operation it retained is exactly as inefficient as before. Only the intensity metrics surface what actually happened.

The evidence that renewables are not yet winning this fight is direct: fossil-fuel energy rose 17% in 2025 and made up 71% of the sector’s total energy. Green power is being added, but it is being outrun by the diesel-heavy demands of deeper, lower-grade operations.

A gold producer reporting lower Scope 2 emissions this year may simultaneously be burning far more energy per ounce, and those two facts do not contradict each other. They are features of the ore body being mined, not signs of management success or failure. Separating the two is essential before you compare any producers on ESG grounds.

South Africa’s deep-level mines and the Eskom problem

The abstract mechanisms become concrete the moment you look at South Africa. Deep-level underground mines face an energy problem that open-pit and surface operations elsewhere never encounter, and it has no simple technological fix.

Depth is the driver. The deeper a mine goes, the more energy it takes to move rock and people to the surface, push air down to keep workers alive, cool an environment that grows hotter with every metre, and pump out water that constantly seeps in. Four demands scale relentlessly with depth:

Sibanye-Stillwater and Harmony Gold top the energy-intensity rankings in the Metals Focus dataset, and the reason is only partly their own operational choices. The bigger factor is the grid they are plugged into.

Eskom, South Africa’s state power utility, generated roughly 82% of its electricity from coal according to the Metals Focus Gold ESG Focus 2026 report. That means the Scope 2 emissions of any heavy Eskom user are structurally inflated by the grid itself, before a single decision is made underground.

Sibanye-Stillwater and Harmony Gold are “very reliant” on Eskom’s coal-heavy electricity, “which adds a significant amount to their emissions,” said Sarah Tomlinson of Metals Focus.

This complicates any cross-jurisdictional comparison. A South African producer’s Scope 2 intensity can look poor beside a Canadian or West African peer not because its management is weaker, but because it is drawing power from a coal-dominated grid. Worth noting that the earlier Metals Focus 2025 report put Eskom at around 70% coal, as reported by Mining Magazine in September 2025; the difference from the 82% figure reflects different reporting periods and methodology rather than a single settled number.

For an investor, the implication is direct. Penalising a South African deep-level producer on raw Scope 2 intensity, without adjusting for grid carbon content, risks reading a structural grid constraint as an operational failure and systematically underrating management that may in fact be actively cutting intensity.

Solar as a partial offset, not a structural solution

Harmony Gold shows the offset in action. Its intensities are “high, but they’re coming down because the solar plants are coming on stream,” Tomlinson noted. On-site solar reduces the volume of coal-heavy Eskom power a mine has to buy, which directly cuts Scope 2 emissions even in a deep-level environment.

There is a benchmark for how far this can go. B2Gold’s Otjikoto mine draws roughly 79% of its electricity from renewables, showing the penetration achievable at surface or near-surface operations where solar has room to work.

The limits are just as clear. On-site solar addresses grid electricity, but it does nothing about the fundamental energy demand created by depth, and it cannot touch the diesel that made up 71% of the sector’s total energy use in 2025. Solar trims the Eskom exposure at the margin. It does not solve the physics of mining a kilometre underground.

Equipment electrification addresses the diesel problem directly, targeting the 71% fossil-fuel share that solar cannot reach, though the capital requirements and operational disruption of transitioning underground fleets mean payback timelines vary substantially depending on mine depth and production profile.

How producers compare, and what genuine efficiency progress looks like

Sector averages hide the real story, which is how much company-level performance diverges. The spread runs from Centerra Gold at 0.37 tCO2e per ounce to an industry average near 0.9 tCO2e per ounce, a gap of more than two times. That range is proof that asset mix and management decisions genuinely move the needle.

B2Gold’s 2025 result illustrates the most direct lever of all, and it is one investors routinely misread. Its energy intensity fell from 11.7 GJ per ounce in 2024 to 9.7 GJ per ounce in 2025. The primary reason was not that the company slashed energy use; it was that gold production rose, spreading a similar energy load across more ounces.

Eldorado Gold offers a cleaner decarbonisation signal, combining a low starting point of 0.42 tCO2e per ounce with 46% progress toward its greenhouse gas mitigation target and 76% of its water recycled. Endeavour Mining describes itself as a low-intensity producer at 0.57 tCO2e per ounce, though that is self-reported positioning rather than a Metals Focus benchmark figure.

A dramatic one-year improvement in intensity can come from producing more gold rather than from any decarbonisation action, which means intensity metrics are as sensitive to production trajectory as they are to genuine efficiency work.

That is why single-year figures are so easy to misread. A drop in energy intensity that lines up with a production ramp tells a very different story than the same drop achieved while output stayed flat. Look at intensity trends across several years, and always alongside the production numbers driving them, before drawing any ESG conclusion.

Reading the ESG scoreboard before the next resource cycle

The core insight is that no single number settles the question. Absolute emissions, per-ounce intensity, production trajectory, grid carbon content, and asset-portfolio changes are five distinct signals, and any ESG conclusion drawn from fewer than all five is incomplete.

The three-year climb in emissions intensity across 2023, 2024, and 2025, together with energy intensity now 47% above its 2016 baseline, confirms this is a structural direction and not a one-year blip.

The sharpest risk to watch is ESG arbitrage. A producer that divests its high-emitting mines will post falling absolute emissions with zero improvement in the efficiency of what it kept, and intensity is the only metric that exposes it. Contrast that with B2Gold, where production growth genuinely pulled intensity down. One is accounting improvement; the other is structural.

Run any gold producer’s ESG claim through this checklist:

Remember that ore-grade decline and mine depth are geological facts, not management failings. An evaluation framework that ignores them will misallocate capital. The sector’s trajectory through the next resource cycle will be shaped more by ore grades, mine depth, and grid decarbonisation than by any renewable pledge announced today, and investors who grasp that start from a materially stronger position.

For readers wanting to trace the longer-arc trajectory beyond the 2025 snapshot, our full explainer on mining emissions and decarbonisation pathways examines the structural interventions, grid transition timelines, and technology adoption curves that determine whether the sector can reverse the intensity trend across the next resource cycle.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. Financial projections and forward-looking assessments are subject to market conditions and various risk factors.