Mountain Pass Cutoff Math: Why 6% REO Decides the 2025 Restart

The arithmetic of industrial scale rarely bends to political narratives. When analysts dissect the Mountain Pass recovery, they fixate on defense contracts and trade barriers, yet the true determinant sits quietly in a single metallurgical parameter: 6%. This threshold is not a policy target or a market forecast; it is the hard boundary where extraction transitions from theoretical resource to bankable concentrate. Every ton moved through the facility must clear this line to justify energy, labor, and chemical inputs.

Historical production cycles reveal that grade distribution consistently overrides macroeconomic headlines. Tariff adjustments shift short-term margins, but they cannot manufacture higher concentrations from low-grade feedstock. The deposit’s internal geometry does the heavy lifting, dictating whether downstream separation remains economically viable. When reserve grades hover just above the cutoff, the entire supply chain stabilizes without requiring external subsidies or regulatory exemptions.

Recognizing this constraint reframes how we evaluate critical mineral infrastructure. Capital deployments and strategic partnerships matter, yet they function as accelerants rather than foundations. The actual feasibility of scaling operations rests on whether mined material sustains the required concentration through every processing stage. Ignoring this fundamental metric leads to flawed projections about industry timelines and technological breakthroughs.

Cutoff Math

At Mountain Pass, the 6% REO threshold is not an arbitrary target; it is the mathematical boundary where revenue from recovered rare earth oxides exactly equals the combined cost of mining, hauling, crushing, and flotation per tonne. Below this line, every tonne milled destroys margin by consuming energy and reagents to produce less value than the input costs; above it, every tonne adds margin. This breakeven mechanism dictates that the restart's viability hinges entirely on mill feed grade, not headline production volume. The deposit's geology enables this precision: the ore consists of bastnaesite, a fluoro-carbonate REE mineral hosted in carbonatite, which floats cleanly at coarse grind sizes. This mineralogical advantage allows a high-tonnage, grade-sensitive mill circuit to operate efficiently, meaning the cutoff can be enforced directly at the pit face rather than relying solely on downstream lab sorting.

The binding constraint for 2025 shifted dramatically following the April 2025 halt of concentrate shipments to China, triggered by tariff escalation and Chinese export controls. With external sales blocked, MP was forced to process domestically, making internal mill-feed grade the sole determinant of throughput economics rather than saleable concentrate tonnage. This operational pivot exposes the downstream dependency: MP's 2024 NdPr oxide output highlights a severe bottleneck relative to stated annual targets. Separation recovery cannot manufacture NdPr that was never present in the mill feed; thus, any slippage in upstream grade directly throttles downstream oxide production, rendering the 6% cutoff the critical lever for meeting capacity targets.

Mountain Pass's ~6.4% reserve grade occupies a precarious position within the global deposit landscape. Comparative data from published technical reports places Mt Weld's central lanthanide deposit at roughly 11–12% REO, Nolans at ~4.3%, Bear Lodge at ~1.6%, and Round Top at ~0.6%. This establishes that Mountain Pass sits in a narrow band where a 6% cutoff is both feasible and essential; deposits like Nolans or Bear Lodge lack the headroom for such strict cut-offs without destroying mine life, while higher-grade assets like Mt Weld tolerate wider margins. However, the effective breakeven cutoff is higher than naive revenue-minus-cost math suggests. Published bastnaesite flotation studies and MP's own disclosures indicate concentrate recoveries in the ~60–70% range, meaning roughly a third of contained REO in feed above cutoff is lost to tailings. This recovery penalty raises the true economic threshold, demanding rigorous grade control to ensure the mill feed sustains sufficient metal content to offset processing losses.

The Evidence

When evaluating the restart, ignore headline concentrate tonnage. The canonical metric is cost per kilogram of NdPr produced at ≥6% REO mill feed. Only the 6% cutoff delivers that metric without sacrificing the throughput required to meet the NdPr commitment. Everything else is accounting theater.

Geostatistical estimation at Mountain Pass carries structural blind spots that headline production metrics obscure. The canonical decision rule—cost per kilogram of NdPr at ≥6% REO mill feed—assumes a deterministic relationship between cutoff selection and realized yield, but resource models inherently smooth heterogeneity. When you drill into the variography of the Mountain Pass deposit, the spatial continuity of high-grade lenses degrades rapidly beyond short-range sampling intervals. This means that while bulk tonnage estimates may suggest adequate reserves, local grade variability can cause significant deviation from the expected recovery curve. The data does not capture the stochastic nature of block model realization; it reports averages that mask the risk of localized dilution spikes. Consequently, relying solely on aggregate reserve grades without accounting for conditional simulation variance introduces a systematic bias toward overestimating recoverable NdPr output.

Variance across cases emerges when comparing different mining faces or pit phases. Not all ore zones behave identically under milling constraints. Some domains exhibit tighter grade control due to consistent mineralization styles, while others display erratic REE distribution driven by hydrothermal alteration patterns. This heterogeneity affects the efficiency of the flotation circuit and the subsequent separation yields. In domains with complex gangue associations, the cost structure shifts non-linearly as cutoff grades approach the threshold. The marginal cost of processing lower-grade material rises faster than linear projections suggest because liberation characteristics change with mineralogy. Therefore, the cost-per-kg metric must be evaluated against domain-specific performance data rather than plant-wide averages. Ignoring this variance leads to suboptimal cutoff decisions that erode margins even when overall feed grades appear compliant.

Three Cutoffs, One Winner

The rule breaks under specific geomechanical and processing anomalies. If the mill feed contains elevated levels of deleterious elements such as thorium or uranium, the processing costs escalate disproportionately, invalidating the standard cost model. Similarly, if the ore exhibits significant textural complexity where rare earth minerals are intimately associated with silicate gangue, the recovery efficiency drops below baseline assumptions. These conditions create scenarios where maintaining a 6% REO cutoff becomes economically unviable despite favorable NdPr prices. Additionally, equipment downtime or reagent supply disruptions can alter the cost base dynamically, rendering static cutoff calculations obsolete. Operators must monitor these process variables in real-time and adjust cutoff strategies accordingly. Failure to do so results in operating losses even when the primary thesis regarding grade thresholds holds true under ideal conditions.

Geostatistical resource estimation at Mountain Pass carries structural blind spots that headline production metrics obscure. The canonical decision rule—cost per kilogram of NdPr at ≥6% REO mill feed—assumes a uniform feed, but the carbonatite’s short-range grade variability and pit-contact dilution routinely pull run-of-mine material 0.3–0.5 percentage points below the reported 6.4% reserve grade. Because MP Materials does not publish a formal cutoff grade in its public filings, analysts are inferring the constraint rather than reading it directly from engineering reports.

The basket composition compounds this sensitivity. While NdPr anchors the price floor, lanthanum and cerium dominate the output stream. When La/Ce credits collapse toward zero, the effective breakeven cutoff rises well above 6%, and the whole grade argument tightens. A recovery drop at constant 6% feed raises unit cost per kg NdPr by roughly 15%, enough to erase the 6%-cutoff margin. This is not a theoretical edge case; it is the mechanical reality of bastnaesite flotation when gangue mineralogy shifts or oxidation states vary across ore zones.

What the Data Doesn't Tell You

Rule 1 — Track mill feed grade, not mine output. Quarterly disclosures that highlight rising concentrate tonnes while quietly showing contained REO per tonne milled slipping below ~6% signal deteriorating economics. Higher throughput at lower grade is dilution, not growth. Verify the actual head-grade reported in processing logs, not just shipped volume.

Rule 2 — Convert everything to cost per kg NdPr. Ignore REO-in-concentrate headlines. The only metric testable against the price floor is fully-integrated cost per kilogram of separated NdPr. Any analysis stopping at concentrate misses downstream separation losses, solvent extraction overhead, and waste treatment. If the math doesn’t resolve to a per-kg separated product figure, it’s incomplete.

Rule 3 — Price the basket honestly. Stress-test the cutoff assuming La and Ce credits sit at or near zero. Heavy rare earth baskets often subsidize light rare earth economics in bull markets, but when credit values compress, the effective breakeven cutoff rises above 6%. If the case only works with meaningful La/Ce revenue, the margin is thinner than reported and highly sensitive to demand shifts.

What the Grade Curve Hides

Rule 4 — Demand recovery-adjusted cutoff math. A feed grade is only economic after applying a 60–70% flotation recovery band and an ~85% NdPr separation recovery. Computing cutoffs on contained REO alone understates the true breakeven by roughly a third. You must model the full conversion chain: ore to concentrate, concentrate to mixed oxides, mixed oxides to separated NdPr. Only then does the grade threshold reflect reality.

Rule 5 — Watch the stockpile ledger as the honesty check. A disciplined 6% cutoff banks sub-grade material into separate stockpiles. A falling average feed grade paired with rising total tonnes moved signals cutoff erosion. This divergence is the single most reliable early warning that the restart’s margin story is being averaged away by low-grade blending.

The mechanism is straightforward: every point of cutoff slippage converts Western world’s highest-grade rare earth ore into margin-destroying dilution. Monitor the head-grade, convert to separated cost, stress-test without heavy basket credits, apply full recovery chains, and watch the stockpile ledger for blending drift. That is how you read the restart.

Worked Case

Start with the mill feed. One million tonnes of ore at 6.4% REO, processed through standard flotation at roughly 65% recovery, yields approximately 41,600 tonnes of contained rare earth oxides in concentrate. That volume aligns with the scale MP Materials reported for its 2024 campaign. From there, the arithmetic bridge to the actual pricing benchmark begins: Mountain Pass concentrate typically carries about 15% NdPr within its total REO basket, and downstream separation recovers roughly 85% of that fraction. Multiplying those factors against the 41,600 t REO baseline produces approximately 5,300 tonnes of neodymium-praseodymium oxide. This is the exact mass stream that the prevailing price floor actually prices, not the bulk REO tonnage that headline reports chase.

Revenue anchors first on that floor. Five thousand three hundred tonnes of NdPr × $110 per kilogram generates roughly $583 million in floored product revenue. Add blended lanthanum-cerium credits—typically valued around $1.50 per kilogram applied to an estimated 35,000 tonnes of La+Ce oxides recovered alongside the heavy fractions—and the total lands near $636 million. Spot volatility never enters this calculation because the floor removes it; the case rests entirely on whether throughput cost per kilogram of NdPr stays beneath that ceiling.

The cost side requires a specific framing discipline. A fully integrated operating expense of roughly $9,000 per tonne of ore milled (covering pit haulage, crushing, grinding, flotation, solvent extraction, and refining) might tempt a naive division by the 5.3 kg of NdPr contained in each tonne of feed, which would falsely suggest a unit cost near $1,700/kg. That math ignores two structural realities: the plant processes the entire REO basket simultaneously, and La/Ce credits offset a meaningful slice of the shared conversion overhead. When you allocate total-REO throughput costs across the full concentrate stream and net out the light-ear credits, the true unit NdPr cost at 6.4% feed settles in the $60–$75/kg range. Drop the feed to 5.5% REO, however, and the same fixed conversion chain must process more waste rock to hit the same concentrate target. The allocation shifts, pushing unit NdPr cost upward by roughly 14%, which consumes most of the margin between production cost and the price floor.

Run the sensitivity curve to see where the binding constraint actually sits. At 7.0% feed, allocated cost per kilogram of NdPr remains comfortably below $60. At 6.0%, it climbs into the low $70s. At 5.0%, the credit offset shrinks while throughput dilution rises, driving unit cost past $90. Extrapolating the slope shows the cost-per-kg-NdPr curve intersecting the price floor somewhere between 5.2% and 5.4% feed grade. That intersection is why the 6% cutoff functions as an operating discipline rather than a rounding error: every point of slippage below it compresses the margin envelope until the floor itself becomes the only thing keeping the balance sheet intact.

The verdict follows directly from the math. At 6.4% mill feed the restart clears the price floor with measurable margin. At 5.5% it survives, but the buffer evaporates quickly if recovery dips or energy costs tick up. Below roughly 5.3% feed, the allocated cost per kilogram of NdPr breaches the floor regardless of policy support. The cutoff, not the export regime or spot price swings, is the binding constraint on whether the 2025 restart succeeds.

How to Read the Restart

Reading the Mountain Pass restart requires abandoning headline tonnage and focusing exclusively on mill feed grade. The operation’s ~6.4% reserve grade leaves virtually zero tolerance for cutoff slippage. When you strip away marketing language, five operational rules dictate whether the 2026 campaign survives the price floor.

Rule 1 — Track mill feed grade, not mine output. Quarterly disclosures that highlight rising concentrate tonnes while quietly showing contained REO per tonne milled slipping below ~6% signal deteriorating economics. Higher throughput at lower grade is dilution, not growth. Verify the actual head-grade reported in processing logs, not just shipped volume.

Rule 2 — Convert everything to cost per kg NdPr. Ignore REO-in-concentrate headlines. The only metric testable against the price floor is fully-integrated cost per kilogram of separated NdPr. Any analysis stopping at concentrate misses downstream separation losses, solvent extraction overhead, and waste treatment. If the math doesn’t resolve to a per-kg separated product figure, it’s incomplete.

Rule 3 — Price the basket honestly. Stress-test the cutoff assuming La and Ce credits sit at or near zero. Heavy rare earth baskets often subsidize light rare earth economics in bull markets, but when credit values compress, the effective breakeven cutoff rises above 6%. If the case only works with meaningful La/Ce revenue, the margin is thinner than reported and highly sensitive to demand shifts.

Rule 4 — Demand recovery-adjusted cutoff math. A feed grade is only economic after applying a 60–70% flotation recovery band and an ~85% NdPr separation recovery. Computing cutoffs on contained REO alone understates the true breakeven by roughly a third. You must model the full conversion chain: ore to concentrate, concentrate to mixed oxides, mixed oxides to separated NdPr. Only then does the grade threshold reflect reality.

Rule 5 — Watch the stockpile ledger as the honesty check. A disciplined 6% cutoff banks sub-grade material into separate stockpiles. A falling average feed grade paired with rising total tonnes moved signals cutoff erosion. This divergence is the single most reliable early warning that the restart’s margin story is being averaged away by low-grade blending.

The mechanism is straightforward: every point of cutoff slippage converts Western world’s highest-grade rare earth ore into margin-destroying dilution. Monitor the head-grade, convert to separated cost, stress-test without heavy basket credits, apply full recovery chains, and watch the stockpile ledger for blending drift. That is how you read the restart.

What to do next

Frequently Asked Questions

What specific mineralogical property allows Mountain Pass to enforce the 6% cutoff directly at the pit face?

The ore consists of bastnaesite, a fluoro-carbonate REE mineral hosted in carbonatite, which floats cleanly at coarse grind sizes.

How did the April 2025 supply chain disruption change what determines throughput economics at the facility?

With external sales blocked by tariff escalation and Chinese export controls, internal mill-feed grade became the sole determinant of throughput economics rather than saleable concentrate tonnage.

Why does the true economic threshold for profitability exceed the basic 6% revenue-minus-cost calculation?

Published flotation studies indicate concentrate recoveries in the ~60–70% range, meaning roughly a third of contained REO in feed above cutoff is lost to tailings.

Which comparable global deposit lacks the headroom to sustain a 6% cutoff without destroying its mine life?

Nolans sits at ~4.3% REO, making a 6% cutoff policy infeasible because it would exceed the average grade and destroy the resource base.

What processing variables can invalidate the standard cost model even when NdPr prices remain favorable?

Elevated levels of deleterious elements such as thorium or uranium cause processing costs to escalate disproportionately, invalidating the standard cost model.

How does short-range spatial variability in the deposit affect recovery predictions compared to bulk reserve estimates?

Geostatistical models inherently smooth heterogeneity and report averages that mask the risk of localized dilution spikes, introducing a systematic bias toward overestimating recoverable NdPr output.

Quick answers

Also worth reading: Unlock hidden energy with vital trace minerals from the earth: Unlock hidden energy with vital · Ancient woolly mammoth yields the oldest RNA ever found on earth: Ancient woolly mammoth yields the · One rare fossil discovery finally settles the mystery of the Nanotyrannus: One rare fossil discovery finally