SITELIVE What happened
Core Lithium told the ASX on 8 September 2026 that the first spodumene concentrate has been produced from the recommissioned processing plant at its wholly owned Finniss Lithium Operation, 88 km by sealed road from Darwin Port in the Northern Territory. The company describes it as a key milestone in a staged restart: mining resumed at the Grants open pit in May, the crushing circuit was recommissioned in August to build crushed-ore stockpiles, and the dense-media-separation plant has now run ore to product. Core says first concentrate arrived within six months of the Final Investment Decision announced on 18 March and in line with the September-quarter target it set at FID.
The announcement is explicit that the plant “remains in the commissioning and optimisation phase” and that the work ahead is “optimising plant performance, operational reliability and throughput” through ramp-up. First shipment of newly produced concentrate is targeted for the December quarter 2026, with the logistics chain described as already operational. Managing director Paul Brown’s quote frames it as delivered “on schedule and on budget” and credits “the value of the existing infrastructure at site.”
Why the plant work is the practitioner story
Finniss is a brownfield restart of a plant that ran, and struggled, before the operation was paused in mid-2024. Core’s May 2025 Restart Study set out a plan built on the existing plant with targeted upgrades rather than a rebuild, and the 8 September announcement lists what was actually done before recommissioning: screen refurbishments, enhancements to the rolls crusher and tails thickener circuits, improvements to the ferrosilicon distribution system, and consolidation of the crushing and DMS plant operations into a single control room. Core says these upgrades are expected to improve mineral liberation and recovery while increasing throughput capacity by approximately 20% to 1.2 Mt a year.
Each of those items maps to a known DMS pain point. Screening governs the size fractions fed to each DMS stage; a rolls crusher governs how much of the coarse spodumene is liberated without being ground to fines that DMS cannot recover; ferrosilicon distribution governs the medium density that separates spodumene from lighter gangue, and its consistency is a common cause of recovery swings; the tails thickener governs water recovery and whether the back end limits the front. Merging two control rooms is a staffing and communication change as much as a technical one. None of these is exotic. That is the point: the restart plan bet on fixing the plant it had.
The honest limits
The announcement contains no production numbers. It does not report tonnes of concentrate produced, concentrate grade, plant recovery, throughput rate achieved, or how many hours the plant has run since first product. “First concentrate” in a commissioning context can mean a single campaign, and the company’s own language — commissioning, optimisation, “progressively build towards reliable and sustained production” — should be read as a caution that reliable output is not yet demonstrated. The 20% throughput uplift is an expectation stated in the announcement, not a measured result, and the plant’s ability to hold 1.2 Mtpa on Grants ore will only be known from the quarterly reports.
The economics rest on a plan, not on this event. At FID Core said the restart was funded by a US$70 million convertible note from Glencore and InfraVia, a US$50 million senior secured loan from Nebari and an intended A$120 million equity raise, with first ore from the BP33 underground mine expected in mid-2027 and ramp-up to nameplate 1.2 Mtpa in mid-2028. Grants is the near-term feed and early cash flow; the 20-year life and the 214 ktpa concentrate capacity on the project page depend on BP33 working. The lithium price is not addressed in the announcement at all, and first shipment is still a quarter away.
What it means for operators
For anyone running or planning a brownfield restart, Finniss is a live case of the low-capital path: fix the bottlenecks, merge the control rooms, restart on the pit you already have and develop the underground in parallel. The next two quarterly reports will show whether that path delivers the recoveries and throughput the Restart Study assumed. Watch the ferrosilicon consumption and the fines losses, because those are where DMS restarts have historically given back the gains.
For contractors and suppliers in the Top End, a restart that has produced concentrate is a materially different customer from one that has announced FID: the haul from site to Darwin Port resumes, the mining contract at Grants continues and BP33 development is under way. Whether it holds depends on the price and on the plant, in that order.