Sodium-Ion Battery vs LFP for Grid Storage: Chemistry, Cost and Cycle Life

For a decade the sodium-ion battery has been the chemistry that was always two years away. In 2026 that finally stopped being true: CATL announced a sodium-ion containerised storage system with first deliveries in China scheduled for September, while the best-funded Western sodium-ion startup, Natron Energy, shut its doors a year earlier despite shipping product. Both facts are real, and they point at the same lesson. A sodium-ion battery is now technically credible for stationary storage, but its commercial case is a question of manufacturing scale and lithium price, not of physics alone.

That matters because lithium iron phosphate (LFP) is the incumbent that sets the bar. BloombergNEF put stationary-storage battery packs at $70/kWh in 2025, the cheapest segment of the whole lithium-ion market. A challenger has to beat that number, or beat LFP on something a spreadsheet of $/kWh does not capture: cold-weather behaviour, cycle life, or exposure to a volatile lithium carbonate market.

You will leave with a working model of how the two cells differ at the electrode level, which published numbers are trustworthy and which are vendor claims, a cost-driver analysis, and a decision framework for when to pilot sodium-ion and when to keep buying LFP.

What this covers: the electrochemistry, energy density and cycle-life evidence, safety, the cost stack, the 2025-2026 commercial state of play (CATL, HiNa, Peak Energy, Natron), failure modes, and a practical recommendation.

Context and Background: Why Stationary Storage Reopened the Chemistry Question

Lithium-ion cells won grid storage almost by accident. They were optimised for phones, then electric vehicles, and stationary operators inherited the cost curve those markets paid for. Within lithium-ion, LFP took over stationary duty because it trades energy density for cycle life and thermal stability, and a container on a substation pad does not care about the weight of its cells. BNEF’s December 2025 survey reported an average lithium-ion pack price of $108/kWh, with LFP packs across all segments averaging $81/kWh and stationary storage lowest at $70/kWh (BloombergNEF). Those are pack prices, not installed system prices, and they reflect Chinese manufacturing; the same survey put North America roughly 44% and Europe roughly 56% above China.

The pressure on that picture is lithium itself. Chinese lithium carbonate prices rose about 47.7% during 2025 according to Caixin’s reporting, and BMI has described 2026 spot prices as well above its fundamentals-based view, forecasting a 2026 average near $20,100 per tonne against about $10,502 in 2025 (Mining Reporters on BMI). BMI also lists sodium-ion adoption as a downside risk to lithium demand. A material that doubles in price within a year is exactly the exposure that makes a procurement team ask whether a second chemistry is worth qualifying.

Sodium offers three structural arguments. It is abundant and not geographically concentrated the way lithium brine and spodumene are. Its carbonate precursor is reported to cost about 1% of lithium carbonate. And the chemistry avoids cobalt, nickel and copper on the anode side. What it gives up is energy density, which for a stationary container is a smaller penalty than for a car.

The battery-supply-chain side of this story connects to other topics we cover. Manufacturing quality and traceability for a new chemistry are the subject of our battery gigafactory digital twin reference architecture, and the regulatory data burden that follows any cell sold in Europe is covered in battery passport and PLM under EU regulation. Demand for storage is also being pulled by data centres, discussed in our analysis of AI data-centre power, grid and nuclear.

How a Sodium-Ion Battery Differs from LFP at the Electrode Level

At the cell level a sodium-ion battery works exactly like a lithium-ion battery: a rocking-chair mechanism in which Na+ ions shuttle between a cathode and an anode through a liquid carbonate electrolyte while electrons take the external circuit. The differences that matter are the ion size, the anode host material, and the current collector.

Figure 1: Rocking-chair architecture of a sodium-ion cell versus an LFP cell. The mechanism is identical; the anode host and current collector differ.

Figure 1 shows the two cells side by side. On charge, ions leave the cathode, cross the electrolyte and insert into the anode; on discharge the flow reverses. A separator keeps the electrodes apart while letting ions pass. Because the topology is the same, sodium-ion cells can be built on lithium-ion production lines with modest changes, which is the main reason the industry thinks it can scale faster than a truly new chemistry such as solid-state. CATL describes its stationary product as fully compatible in size with lithium batteries (CnEVPost).

The ion is bigger and the anode must change

A sodium ion is larger than a lithium ion; textbook ionic radii are roughly 1.02 angstrom for Na+ against 0.76 angstrom for Li+. Sodium also sits at a less negative standard potential (about -2.71 V versus -3.04 V against the standard hydrogen electrode). Two consequences follow. The cell voltage tends to be lower for a given cathode type, and the ion is harder to host in a tight crystal lattice.

Graphite, the universal lithium-ion anode, stores sodium poorly in conventional carbonate electrolytes. The industry answer is hard carbon, a disordered non-graphitising carbon with a mix of turbostratic domains, nanopores and defects. Sodium stores in hard carbon through adsorption on defects, insertion between graphene-like layers and pore filling, which produces the characteristic sloping and plateau voltage profile. Practical reversible capacities are commonly reported in the low hundreds of mAh/g, comparable to graphite’s 372 mAh/g theoretical figure for lithium, but initial coulombic efficiency and cost are weaker. Hard carbon is therefore the most debated line item in the sodium-ion bill of materials.

Sodium-ion cathodes fall into three families, each with a different trade-off. Layered oxides (sodium transition-metal oxides, often with nickel, iron, manganese) offer the highest energy density and are favoured for vehicles but are more air-sensitive. Prussian blue and Prussian white analogues are low-cost, open-framework materials that allow fast sodium diffusion; Natron built its product on this family. Polyanionic compounds such as sodium iron pyrophosphate phosphate (NFPP) and sodium vanadium phosphate trade energy density for stability and long life, and Peak Energy’s stationary system uses NFPP (ESS News).

Note what this means for any statement beginning “sodium-ion batteries can…”. There is no single sodium-ion chemistry. A layered-oxide cell and an NFPP cell have different energy density, cycle life, voltage curve and cost. When a vendor quotes a number, the first question is which cathode produced it. CATL has not published the full cathode chemistry behind every product, so treat chemistry-level conclusions about its cells as inference.

Aluminium on both electrodes

A frequently cited practical advantage is that sodium does not alloy with aluminium at the anode potentials involved, so inexpensive aluminium foil can serve as the current collector on both sides, replacing the more expensive and heavier copper foil lithium-ion cells need on the anode. This is a cost and weight saving at the foil level. It is also commonly cited as enabling transport at zero volts, because the aluminium does not dissolve when the cell is fully discharged the way copper can. That second claim is widely repeated by sodium-ion developers, but it is a developer claim and we have not found an independent regulatory assessment, so we flag it as vendor-reported.

Voltage and the state-of-charge problem

LFP has a famously flat discharge plateau near 3.2 V per cell. That flatness is great for power delivery and terrible for state-of-charge estimation, because voltage barely moves across most of the usable range. Sodium-ion cells, particularly with hard carbon, show a more sloped curve. This is a genuine operational difference: a sloped curve makes voltage-based state-of-charge estimation easier in principle, but it also means usable energy is more sensitive to the discharge cut-off, and battery management software tuned for an LFP plateau must be re-parameterised rather than just re-labelled.

Energy Density, Cycle Life and Safety: What the Numbers Actually Say

The direct answer for a planner is this: a current sodium-ion battery stores less energy per kilogram and per litre than LFP, claims comparable or better cycle life and low-temperature performance, and has far fewer field-years of evidence. The headline figures below are mostly vendor-reported, so each is attributed.

Energy density

CATL’s Naxtra, announced for passenger vehicles in April 2025, was reported at 175 Wh/kg, with an operating range down to -40 degrees C (Battery Technology Online). Trade coverage of that launch disagrees on timing: one article says mass production began in June 2025, while other headlines say CATL planned mass production of Naxtra packs by the end of 2025. We treat the start date as unresolved and rely only on 175 Wh/kg being CATL’s claim. Northvolt earlier announced a sodium-ion cell validated at 160 Wh/kg; Northvolt later went bankrupt, so that is a laboratory-validated milestone and not a product you can buy.

Mainstream LFP cells sit in a similar range in gravimetric terms (commonly quoted between roughly 160 and 205 Wh/kg depending on format and generation), so the gap in kilograms is narrower than folklore suggests. The more important gap for stationary storage is volumetric: sodium-ion cells are generally less dense per litre because of lower voltage and the open structure of the active materials. That means more cells, more racks and more land per MWh. CATL’s own container data illustrates the issue indirectly: its TENER Sodium unit is described as just under 30 MWh at roughly 42 tons (ESS News), and CATL has not published a like-for-like LFP footprint comparison in the coverage we reviewed.

Cycle life

Cycle life is where the claims are boldest, and where evidence is thinnest. CATL states that TENER Sodium achieves 15,000 cycles at 25 degrees C, with more than 10,000 cycles at 45 degrees C, a service life it equates to 25-30 years. Coverage disagrees on the end-of-life criterion: one outlet quotes 70% state of health, while a secondary analysis quotes 80% capacity retention. We report the discrepancy rather than choosing one, and the criterion changes the meaning of the number a great deal, since a cell reaching 70% takes far more cycles than one required to stay above 80%.

For context, a secondary article compares this with CATL’s 530 Ah LFP cell at 10,000 cycles to 70% state of health. We could not trace that comparison to a primary CATL datasheet, so treat it as indicative. The honest summary is that sodium-ion cycle life is plausibly at least LFP-class, but 15,000 cycles takes decades to validate in the field, and the product shipped for the first time in 2026. Accelerated testing extrapolates; it does not prove.

Natron’s Prussian-blue cells promised 50,000-plus cycles and very fast charging according to reporting at the time of its closure (ESS News), which shows that long cycle life and commercial survival are different things.

Temperature behaviour

Cold performance is the clearest technical differentiator. CATL’s Naxtra claims 90% usable power at -40 degrees C, and TENER Sodium claims over 92% capacity retention at -20 degrees C within a -20 to 45 degrees C operating window. LFP loses substantial capacity and, more critically, cannot be charged at normal rates below freezing without risk of lithium plating on the graphite anode. Sodium-ion cells are reported to be less prone to plating, though hard-carbon sodium plating is still a real concern at high rates. Peak Energy makes the opposite-end argument: its president states that sodium-ion outperforms LFP in cycle life and calendar life at very high temperatures, and its system uses passive cooling to cut auxiliary power by over 90% (ESS News). That is a company claim from a pilot of just over 3 MWh, not a multi-year fleet statistic.

Safety

Safety is the most abused word in this debate. Sodium-ion cells can be fully discharged to zero volts for shipping and storage (vendor-reported), which reduces stored-energy risk during logistics. CATL reports its Naxtra cell passing needle, drilling and sawing tests without smoke or fire, and says TENER Sodium can “suppress fire and explosion under extreme conditions.” Those are vendor test statements; neither amounts to an independent UL 9540A or large-scale fire test result, which is the standard that actually informs fire codes for stationary systems.

LFP is already the safer lithium-ion chemistry because its olivine cathode releases little oxygen on heating. The real question for sodium-ion is not “is it safe” but “does it present a different or lower thermal-runaway profile under standardised abuse tests”, and public data is scarce. Organic carbonate electrolytes are still flammable in both chemistries, so the electrolyte, not the cathode, is often the dominant fuel in a cell fire.

Cost Drivers: What Actually Sets the Price per kWh

Cost is the argument sodium-ion has to win eventually, and it is also the argument where public numbers are weakest. Be careful with any claim of “sodium is X% cheaper”: today it is not, in most markets, and the interesting question is what would have to be true for it to become so.

Figure 2: The cost stack. Lithium price pressure and hard-carbon cost move in opposite directions, and scale decides whether sodium’s cheaper raw materials ever reach the customer.

Figure 2 maps the pieces. Raw material cost is only part of cell cost; plant depreciation, yield loss, electrolyte and separator, and the balance of system (containers, power conversion, installation) make up the rest. A cheap precursor helps only if it is a large enough share of the bill and if the factory is as efficient as the incumbent’s.

The lithium price is the swing variable

Lithium carbonate is a minority of LFP cell cost at low prices and a larger share at high prices. When BMI’s cited spot levels sit near $22,500-$22,900 per tonne, close to double 2025’s average, lithium becomes a meaningful driver of LFP cell cost; at $10,000 it is a footnote. This is why sodium-ion’s competitiveness is cyclical: in the 2023-2025 low-price period, the commercial case collapsed, and in 2026 it revived. A project financed on the assumption that lithium stays expensive carries the same risk as one assuming it stays cheap.

BMI itself expects the surplus to return, forecasting declines toward $14,500-$16,500 per tonne by 2027-2028. If that is right, sodium-ion’s raw-material advantage narrows again just as its factories ramp. Sodium-ion buyers are partly buying a hedge, and a hedge is worth what the volatility is worth.

Hard carbon is the new bottleneck

In sodium-ion, the anode takes the role cobalt or lithium played elsewhere: it is the expensive, immature supply chain. A Chinese report summarised by CarNewsChina quotes Wanhua Chemical’s battery-division manager projecting hard-carbon cost falling from 60,000-70,000 yuan per tonne in 2024 to 35,000-40,000 yuan by 2026, with a long-term target below 25,000 yuan (CarNewsChina). Those are supplier projections, not market prices, and graphite for LFP remains much cheaper per tonne. For a sodium-ion cell to win on cost, hard carbon must fall by roughly half and anode yields must improve.

The same report quotes CATL’s chief technology officer saying the company aims for sodium-ion cell cost to match LFP by the end of 2026 and for total energy-storage-system cost to align with LFP in 2027. Note the wording: a target, set by the vendor, for parity, not for undercutting.

What the market prices say today

One secondary analysis reports sodium-ion cells at about $59/kWh versus about $52/kWh for LFP, citing an industry news site, along with a 26.9% lifecycle-cost advantage per discharged kWh for sodium. We cannot verify either number against a primary price survey, and the lifecycle figure depends entirely on assuming the 15,000-cycle claim holds. Treat both as illustrative of the direction CATL’s marketing is aiming, not as procurement inputs.

Here is a worked illustration, with assumed inputs, of why cycle life changes the answer. Suppose an LFP cell costs $52/kWh and delivers 10,000 cycles at 90% average depth of discharge, and a sodium-ion cell costs $59/kWh and delivers 15,000. Cell cost per delivered kWh-throughput is then about $52 / (10,000 x 0.9) = $0.0058 for LFP and $59 / (15,000 x 0.9) = $0.0044 for sodium-ion, a gap of roughly 25%. Change sodium’s cycle life to 8,000 cycles and it flips: $59 / 7,200 = $0.0082, 40% worse than LFP. The lesson is that the entire economic case rests on a durability number that nobody has yet seen in a decade-long field record. These inputs are illustrative, not measured.

There is also a missing-piece trap. Cell cost is perhaps half or less of installed system cost; the rest is balance of system. If sodium-ion needs more containers per MWh because of lower volumetric density, those costs rise even as cell cost falls. CATL’s own roadmap acknowledges this by targeting system parity a year after cell parity.

Digital traceability costs money too

A new chemistry arrives with new data obligations. The EU Battery Regulation requires battery passports for industrial batteries above 2 kWh from February 2027, covering composition, carbon footprint and recycled content, and sodium-ion cells inherit that requirement just as lithium cells do. Manufacturers with a mature digital thread, of the sort described in our gigafactory digital twin architecture, have an advantage in ramping a new chemistry because quality data from early production can feed process control immediately. That is a cost advantage that does not show up in a materials table.

The 2025-2026 Commercial Scorecard

The commercial picture is uneven, and a fair comparison needs to name who actually delivered what.

HiNa and Datang, China (2024). The largest operating sodium-ion storage project at the time of its commissioning was Datang’s 50 MW / 100 MWh first phase in Qianjiang, Hubei, connected to the grid on 30 June 2024 with 185 Ah sodium-ion cells from HiNa Battery, per pv magazine (pv magazine). It is part of a planned 100 MW / 200 MWh project and surpassed the earlier 10 MWh system at Nanning. It remains the best operational evidence at scale, but a project of this size over two years is not a fleet history.

Peak Energy, United States (2025). A pilot of just over 3 MWh at the Solar Technology Acceleration Center in Aurora, Colorado, built on NFPP cells and described by the company as the largest such system in the world, with nine IPPs and utilities participating and four additional projects contracted for 2026-2027. Its claims (20% lifetime cost reduction, 33% lower degradation over 20 years, up to $75/kWh net-present-value savings) are modelled company projections.

Natron Energy, United States (collapse). Natron ceased operations on 3 September 2025 after announcing the decision on 27 August; its board said efforts to raise sufficient new funding were unsuccessful, and two sites were closed with 95 job losses. A year earlier it had announced a $1.4 billion plan for a 24 GW factory in North Carolina. The failure was financial and not electrochemical, and it is the clearest warning in the sector: a technically promising Prussian-blue cell without lithium-class scale could not survive.

CATL TENER Sodium (2026). CATL unveiled the system at Intersolar Europe in June 2026 as the first field-validated sodium-ion BESS, with first China deliveries in September 2026, around 1 GWh expected to ship by end-2026, and international deliveries from June 2027, per CnEVPost. Reported features include support for one- to eight-hour durations, a round-trip efficiency improvement of about 2% from bidirectional DC voltage control, and fault isolation within 200 milliseconds. We could not find an independent third-party validation of the field-performance claim, and the 1 GWh figure is a modest start next to LFP volumes, which are measured in hundreds of GWh.

Taken together the scorecard shows a technology transitioning from pilots to first commercial deliveries, dominated by one Chinese supplier with an LFP business large enough to cross-subsidise early sodium production. That market structure matters more to a buyer outside China than any single spec.

Where Each Chemistry Wins: A Decision Framework for Grid Storage

The useful question is not which chemistry is better, but which is better for a specific site, duty cycle and risk appetite. The framework in Figure 3 encodes the logic we would use for a first sodium-ion decision.

Figure 3: A site-level decision tree. LFP is the default; sodium-ion earns a pilot when climate, lithium-price exposure or supply diversification outweighs density and bankability.

The tree is deliberately LFP-biased, because the incumbent has a decade of field data, bankable warranties, mature insurers and lenders, and the cheapest verified pack price. Sodium-ion enters only when a specific condition tilts the balance. We unpack each branch below.

Space and weight are cheap in some places and costly in others

Utility-scale storage on rural land rarely suffers from a lower volumetric density; a few extra containers are a small line item. In dense urban substations, behind-the-meter commercial sites, offshore platforms and ships, footprint is the binding constraint, and LFP’s density advantage matters. Sodium-ion’s lighter anode foil and cell mass advantage matters most where mass is penalised, such as transport; for fixed storage it is mostly irrelevant.

Climate extremes favour sodium on paper

A site that sees sustained temperatures below about -20 degrees C is the cleanest sodium use case. LFP cells must be heated before charging in deep cold, and heating consumes auxiliary energy and complicates thermal design. If CATL’s claim of more than 92% capacity retention at -20 degrees C holds in field operation, sodium-ion saves both capacity and parasitic load. The reverse argument, from Peak Energy, concerns hot climates where passive cooling reduces HVAC consumption; it is plausible but rests on a short pilot.

Neither extreme matters in temperate climates where HVAC-managed containers keep LFP in its comfort zone. In that case sodium-ion has to win on price or diversification.

Duty cycle and duration

Cycle-life claims matter most for high-utilisation assets: frequency regulation, daily arbitrage and solar shifting at one to two cycles per day. A site cycling once a day uses 365 cycles a year, so 10,000 cycles is nearly 27 years and 15,000 is over 41 years of nominal life. That is longer than the project, the inverter and the interconnection agreement. In practice, calendar ageing, not cycle count, becomes the binding limit, and the sodium-ion advantage in cycle life is worth little to a once-a-day asset. It is worth more to a site doing multiple daily cycles for ancillary services, where cycle count is consumed quickly.

This is an under-appreciated point in the marketing: for most storage, extra cycles beyond what the project will ever use are unredeemable. The value of a 15,000-cycle claim is in lower degradation per cycle, which reduces augmentation spend, not in an extended lifetime.

Decision matrix

Criterion	LFP	Sodium-ion	Edge
Verified pack price	$70/kWh stationary pack average in 2025 (BNEF)	No independent survey; vendor targets parity by end-2026	LFP
Gravimetric energy density	Roughly 160-205 Wh/kg depending on format	CATL reports 175 Wh/kg for Naxtra; Northvolt validated 160 Wh/kg before bankruptcy	Near tie
Volumetric density	Higher	Lower, more containers per MWh	LFP
Cycle life claims	10,000 to 70% SoH quoted for a CATL LFP cell (secondary source)	15,000 claimed, criterion disputed between 70% and 80%	Sodium on paper
Cold-weather performance	Charging restricted below freezing	Over 92% capacity at -20 degrees C claimed	Sodium
Field history at scale	Hundreds of GWh deployed	Tens to hundreds of MWh in operation	LFP
Raw-material exposure	Lithium price swings	Hard carbon and cathode precursors, sodium carbonate cheap	Sodium
Supplier diversity	Dozens of tier-one producers	One dominant supplier, several smaller Chinese firms, Western startups in distress	LFP
Bankability and insurance	Mature	Immature	LFP

The matrix makes the real structure plain: sodium wins on a few technical axes and on commodity hedging, and loses on every axis that depends on time and scale. The first group is addressable by engineering; the second is addressable only by waiting.

Failure Modes and What Goes Wrong with Sodium-Ion

Before the recommendations, it is worth taking the failure modes seriously. Figure 4 lists the five we think matter most and the mitigation that cuts across them.

Figure 4: Five failure modes for a sodium-ion storage project and the shared mitigation of piloting a tranche with dual sourcing.

Supply is thin. CATL’s initial shipments of about 1 GWh by end-2026 are a rounding error in a global market, and international deliveries start only in mid-2027. A buyer outside China competes with the vendor’s domestic customers and may pay a premium or wait.

The orphaned warranty. Natron’s closure is the template. A 30-year warranty, which CATL assigns to its system, is only as strong as the issuer. For a startup supplier, ask for an escrow, an insurer-backed warranty or step-in rights; for CATL, assess geopolitical and tariff exposure instead.

Density costs real money. Each extra container needs a pad, cabling, fire-code clearances and a larger point of interconnection study. If the site was designed around LFP’s footprint, retrofitting sodium is not free.

The voltage curve bites the BMS. A sloped curve is not a drop-in replacement. State-of-charge estimators, balancing thresholds and protection limits tuned for a flat LFP plateau must be redone. Poorly parameterised BMS software is a classic way to waste a chemistry’s cycle life. Dispatch software that assumes a flat curve can also under- or over-estimate available energy near the knee.

Unproven durability. Fifteen thousand cycles at 25 degrees C is an accelerated-test claim. Hard-carbon anodes have known aging mechanisms, including electrolyte decomposition, sodium plating at high rates and cathode transition-metal dissolution in some layered oxides, and the interaction between them over 15 years is unobserved. The 10,000 cycles reported at 45 degrees C already shows a one-third reduction from the 25 degrees C figure, so temperature management still matters.

There are also three subtler traps. First, vendor tests of fire behaviour are not a substitute for UL 9540A or equivalent large-scale testing, and local fire authorities may not yet have approved sodium-ion systems; permitting is a schedule risk. Second, recycling infrastructure is minimal, and EU recovery-efficiency obligations apply regardless of chemistry. Third, a cheaper chemistry can encourage over-sizing; if cells get cheaper but the interconnection queue and inverters do not, the saving may never reach the project IRR.

Practical Recommendations

For most developers in 2026, the sensible stance is LFP by default, sodium-ion by experiment. Procure the main block of capacity as LFP with a bankable warranty, and reserve a small tranche, perhaps a few percent of the portfolio, for a sodium-ion pilot that generates your own site data. A pilot answers the questions that datasheets cannot: real round-trip efficiency, auxiliary load in your climate, BMS behaviour and permitting friction.

Structure the contract so the vendor, not you, carries the durability risk. Ask for a throughput guarantee in MWh, not just a cycle count, and insist on a defined end-of-life criterion; the 70% versus 80% ambiguity above could be worth years of life. Demand the test report behind the cycle number, including the temperature, depth of discharge and charge rate used. If the supplier is not a tier-one manufacturer, require escrow, insurer-backed warranty or a second-source arrangement.

Model the economics with ranges. Run lithium at $10,000, $15,000 and $22,000 per tonne, cycle life at 6,000, 10,000 and 15,000, and hard carbon at today’s and target prices. If sodium-ion wins only at the optimistic corner of every range, you are buying a hedge and a learning opportunity, not savings. Finally, invest in data: log every cell-level parameter during the pilot, because that dataset is the thing your organisation will keep when vendor claims turn out different.

Default to LFP for the bulk capacity; pilot sodium-ion in a small tranche.
Prioritise cold-climate or lithium-exposed sites for the pilot.
Contract for MWh throughput and a defined end-of-life criterion.
Verify fire testing (UL 9540A or equivalent) and local permitting before ordering.
Re-parameterise BMS and dispatch software; never copy LFP settings.
Require a bankable warranty issuer, escrow or second source.
Revisit the decision when independent field data from the first deliveries appears in 2027.

This article is a technology and systems analysis for educational purposes and is not investment, financial or engineering-certification advice.

Frequently Asked Questions

Is a sodium-ion battery cheaper than LFP in 2026?

Not on independently verified numbers. BloombergNEF’s 2025 survey put stationary lithium-ion packs at $70/kWh, and it did not cover sodium-ion. CATL’s chief technology officer has described cell-cost parity with LFP as a target for the end of 2026 and system parity for 2027. Secondary reports cite sodium cells slightly above LFP today. Sodium-ion can still be cheaper per delivered kWh if its claimed cycle life holds.

How many cycles does a sodium-ion battery last?

Vendors claim a lot, and field evidence is thin. CATL states TENER Sodium reaches 15,000 cycles at 25 degrees C and over 10,000 cycles at 45 degrees C; coverage disagrees on whether the end-of-life point is 70% or 80% of capacity. Natron claimed 50,000-plus cycles for Prussian-blue cells before its closure. Treat all of these as accelerated-test results until multi-year fleet data is published.

Is a sodium-ion battery safer than LFP?

Not demonstrably. Both use flammable organic electrolytes, and LFP is already the most thermally stable mainstream lithium-ion chemistry. Sodium-ion developers report zero-volt shipping and passing needle and saw tests, but those are vendor claims. A fair safety comparison needs standardised abuse results, such as UL 9540A or large-scale fire tests, for the specific cell and system, and that public data is still scarce.

Why does sodium-ion use hard carbon instead of graphite?

Sodium ions are larger than lithium ions, and graphite stores them poorly in standard carbonate electrolytes. Hard carbon, a disordered carbon with defects, nanopores and expanded layers, hosts sodium through adsorption, insertion and pore filling. It delivers practical reversible capacity in the low hundreds of mAh/g, but its first-cycle efficiency and cost are weaker than graphite’s. That makes hard carbon a central cost and supply-chain issue.

What happened to Natron Energy?

Natron, a Prussian-blue sodium-ion maker, announced on 27 August 2025 that it would cease operations, and closed on 3 September 2025 after failing to raise new funding. It shut sites in Holland, Michigan and Santa Clara, California, with 95 job losses, a year after announcing a $1.4 billion, 24 GW North Carolina factory plan. The reported cause was financing, not a chemistry failure.

When will sodium-ion batteries be available outside China?

CATL says its TENER Sodium system will begin deliveries in China in September 2026 and ship internationally from June 2027, with about 1 GWh expected by the end of 2026. Peak Energy, a US company, has four commercial projects contracted for 2026-2027 after its Colorado pilot. Availability, pricing and warranty terms for non-Chinese buyers remain unconfirmed, and permitting may add lead time.

Sodium-Ion vs LFP Batteries for Grid Storage: Chemistry, Cost and Cycle Life