← Back to hub
The Downside

Every way this can fail — and what’s still standing when it does.

A pitch that only describes success is a pitch you shouldn’t trust. So here is the opposite: the map of every point where this can break, and exactly what value survives each break.

Fallback ladder · stage-by-stage resilience · bounded-downside map

The conclusion is the whole reason the ask is structured as a $1.8M option: the downside branch never returns zero. Every failure mode has a fallback with a known chemistry and a known cost, and even total battery failure leaves Emerson holding five standalone capabilities, two of them with genuine humanitarian value. This page is the receipt for that claim.

01 · The science fallback ladder

A staircase, with no cliff.

Each rung is a known chemistry with a known cost. If the electrolyte you want doesn’t work, you step down. You don’t fall.

Fallback ladder · rising cost, never a cliff
Rung 1
Chen’s two cells reproduce at grid-relevant C-rate, and the blend holds. The anchor. Chen 2026’s published 120k-cycle results are two single-cation cells, saturated MgCl2 and saturated CaCl2, run separately; the CaCl2 cell is the stronger (48.3 vs 40.8 Wh/kg). Gate 0 replicates in those two electrolytes. The mixed Mg/Ca cell is the Phase-1 bet: the ratio series starts from the two published endpoints and maps the compositions between them. If the blend keeps both endpoints’ life at C/4, the frontier unlocks: run it on already-mixed brine (skip Ca/Mg separation).
Best case
Rung 2
Mixed at grid rate shows ratio-sensitive behavior → Ca-rich. Ca is the better-behaved divalent (higher selectivity, faster kinetics than Mg). A Ca-dominant mixed electrolyte still uses the “keep divalents, reject monovalents” purification, and still avoids the hard Mg-from-Ca separation.
Modest Δ
Rung 3
Ca-rich underperforms → pure calcium. The stronger published cell (Chen’s CaCl2 cell: 120,000 cycles at 20 A/g, 48.3 Wh/kg, 1.2 V), so the open question is transfer to C/4. Phase 1 tests pure-Ca as a first-class parallel outcome. Feedstock is TETRA / OxyChem battery-grade domestic supply, Mg-free by source (so no separation needed at all). Simpler, lower-risk, ships from purchased CaCl2.
Moderate Δ
Rung 4
Pure Ca fails → pure magnesium. The single-cation Mg version, Chen’s other published cell (40.8 Wh/kg, 1.1 V). Also unproven at Phase-1 rates and pricier (adds Ca/Mg separation ~$400–600/ton (an estimate this site has not yet sourced) on incumbent-supply feedstock), but Mg supply chain is domestic-adjacent and the chemistry is well-studied. The fallback ceiling.
Expensive fallback
At no rung do we fall off a cliff. Rungs 3 and 4 have Chen’s coin-cell data directly behind them; Rungs 1 and 2 are the blends the parallel ratio series maps between those two published endpoints. The frontier is running the proven-or-mapped cell on already-mixed brine, and it hinges on the blend. Even the ceiling still works.
02 · Stage-by-stage resilience

What survives each Phase-1 gate failing.

Every Phase-1 gate has an exit ramp, each with capability that remains sellable.

Gate 1 · Chen 2026 doesn’t independently reproduce
The single biggest risk, and the reason Phase 1’s core deliverable is to prove or kill it, cheaply, before anyone scales.

If Chen’s COP+CuFe-PBA+MgCl2 cell won’t reproduce at grid-relevant C-rates, the chemistry thesis ends here. Phase 1 is designed to expose this in the first months. Kill fast.

What survives: the purification skid, the poly-mineral recovery platform, the battery-materials QA methodology, and the brine-validated instrumentation all still stand (see Beyond the Battery). ~$1.8M bought a definitive answer plus five sellable capabilities, which is a long way from zero.
Gate 2 · Reproduces, but mixed-valent doesn’t co-cycle
Drop down the fallback ladder to Ca-rich or pure-Ca.

The cell still works; the electrolyte spec narrows. Cost rises modestly; the purification path complicates but stays inside known chemistry. The commercial framing shifts from “we deleted the industry’s hardest step” to “we’re the second lab to reproduce Chen at pilot scale.” Still a strong first-mover position.

What survives: everything. The cell + the platform + the capabilities. The mixed-valent IP moat weakens; every other asset stands.
Gate 3 · Cells work, but Track-B in-house purification can’t hit battery grade
Fall back to Nedmag battery-grade supply; sell the purification skill into pharma / water instead.

If Emerson’s DeltaV+ICP polish step can’t reach battery grade at scale, the cell ships on imported Nedmag salt (a mature commercial supply). The purification skid becomes a pharma / drinking-water product, a different market with lower purity requirements and a real buyer network.

What survives: the whole battery program (on imported feedstock), plus the purification skid re-aimed at pharma / water. Two products instead of one.
Gate 4 · Everything works at bench, fails at scale-up
Scale-up is Phase 2+. No Phase-1 capital was committed to it.

If pouch cells at 1 Ah work but 3 MWh skid geometry doesn’t (current distribution, edge effects, electrode-stack mechanics), that is a Phase-2 discovery on Phase-2 capital, with no Phase-1 loss. The known unknowns are named explicitly in the Storage Landscape section under “What We Have Not Yet Proven.”

What survives: the process IP, the data platform, the impurity-tolerance dataset, and the standalone markets remain. Scale-up is Phase 2+; no Phase-1 capital is at risk from it.
03 · The bounded-downside map

Total battery failure at any gate still leaves Emerson holding five capabilities.

The worst case (Gate 1 fails, chemistry ends) still returns a portfolio of sellable process, instrument, and data capabilities. Two of them are good-for-the-world businesses on their own merits.

Capability 1
Automated high-purity salt skid

Pharma / food / drinking-water remineralization. Global-health product. Nedmag’s existing market.

Capability 2
Poly-mineral brine-recovery platform

Cu / B / Li / K / Ba / Sr recovery. Domestic critical-minerals + circular economy.

Capability 3
Desal-brine remediation business

Turn hypersaline discharge into mineral products. May end up larger than the cell.

Capability 4
Battery-materials QA + data platform

Sellable to every battery chemistry (LFP, sodium-ion, whatever wins). De-risks Emerson vs the “wrong chemistry” question entirely.

Capability 5
Brine-validated instrumentation

Rosemount / Micro Motion proof-point into mining, potash, and desalination markets Emerson already serves.

Two goods for the world
Two of the five, safe drinking water (Capability 1) and cleaner desal brine (Capabilities 2 + 3), are worth doing for their own sake. Even a total battery failure moves the world forward on those axes.
Feedstock-supply resilience · a distinct upstream branch (not a sixth capability)
The five capabilities above answer “what survives if the battery fails.” This one answers a different question: “what if a feedstock supply chain fails while the battery is working fine?” A chemistry that runs on Mg-only, Ca-only, or mixed is one no single feedstock supply chain can halt. If Mg tightens (Nedmag disruption, NM solar-pond weather, Chinese export policy), run Ca-dominant (TETRA / OxyChem domestic supply). If Ca tightens, run Mg-dominant (Nedmag / Intrepid). If both flow freely, run mixed for lowest cost. The parallel Phase-1 ratio series proves all three simultaneously: a science hedge and a supply-chain hedge, free with the same experiment. The lithium / cobalt / vanadium worlds cannot escape single-source bottlenecks. This one is engineered not to have one. Mg is the anchor and the Phase-4 whale’s harvest; Ca is the co-lead cation with independent domestic supply. Neither is orphaned; either can carry the load.
04 · Why this is the option-value proof

This page is the literal evidence behind the ask.

The receipt
$1.8M is an option precisely because the downside branch returns durable, sellable capabilities, no matter which gate closes.

The Financial Outlook on the proposal prices the option at under 1% of the ~$200M clears-Gate-1 revenue floor. That floor doesn’t assume the battery succeeds commercially, but about three-quarters of it is battery hardware, so it does assume the chemistry clears Gate 1. If the chemistry fails outright, the floor is the five non-battery capabilities, in the tens of millions. The gates below that, the ones this page walks, are what backstop the worst case with things a CFO can underwrite instead of hope for.

Judge the ask on this page rather than on the ceiling. Fund the floor; hope for the tiers above it.

One caution

The bounded-downside case is not a reason to be complacent about the science. The mixed-valent bet is real, the reproduction of Chen 2026 is genuinely load-bearing, and the fallback ladder rises in cost at every rung. This page’s job is to show that even the worst outcomes have real floors. It does nothing to soften how much we want the best outcome.

Phase 1’s job is to give the science its fairest chance to succeed. This page is what makes it safe to try.