Battery Technology
Why battery chemistry matters for this site
The safety profile of a battery storage facility depends heavily on its chemistry. Lithium-ion cells can undergo thermal runaway, producing hydrogen fluoride and other toxic gases. Sodium-ion cells in the NFPP formulation have shown markedly lower thermal-runaway potential in controlled abuse testing, though sodium-ion is a family of chemistries and not all of them behave the same way. The difference between these chemistries determines the fire risk, toxic emission profile, cooling requirements, and noise of the facility.
Jupiter Power has stated that battery chemistry “has not yet been selected” but is expected to be “lithium ion (LFP) or sodium ion.” In practice, PSE’s tolling agreement is structured around lithium-ion BESS operations and names Samsung SDI as the battery vendor, and all 47 standalone battery storage proposals PSE received in its 2024 RFP were lithium-ion. A full Environmental Impact Statement with an alternatives analysis would put the chemistry question on the public record and require evaluation of whether a safer chemistry is feasible at this site.
Lithium-ion risks
Thermal runaway is an uncontrolled temperature rise inside a battery cell that feeds itself. Once one cell goes, it can cascade through an entire container. The process generates large volumes of flammable and toxic gas.
Incidents
| Date | Facility | Capacity | What happened |
|---|---|---|---|
| 2017–2019 | South Korea (23+ sites) | Various | 23+ fires, $32M+ losses, 522 ESS units shut down nationwide. Root causes: insufficient battery protection, poor installation, BMS integration gaps |
| Apr 2019 | McMicken, AZ | 2 MW | Explosion injured 4 firefighters seriously, 8+ first responders total |
| May 2024 | Gateway, San Diego | 250 MW | Burned 11 days, EPA-ordered cleanup |
| Jan 2025 | Moss Landing, CA | 300 MW | 1,200–1,500 evacuated, largest EPA lithium-ion cleanup in history |
What comes out of a lithium-ion battery fire
- Hydrogen fluoride (HF): 20–200 mg per Wh of capacity. Causes deep tissue burns and pulmonary edema.
- Heavy metals: After the Moss Landing fire, researchers at San Jose State’s Moss Landing Marine Labs estimated 55,000 pounds of toxic metals entered surrounding wetlands. Nickel concentrations in nearby soil spiked to 15 times pre-fire levels. The EPA invoked Superfund authority to order the cleanup.
- PFAS: A 2025 study found per- and polyfluoroalkyl substances in lithium-ion battery fire soot.
- Carbon monoxide, hydrogen cyanide, sulfur dioxide, and various volatile organic compounds.
Reignition
These fires come back. Undamaged cells in a partially burned facility still hold energy, and that energy can reignite days or weeks later. Gateway reignited repeatedly over 11 days.
Why this site is different
The Snoqualmie Valley is ridge-bounded. Temperature inversions in western Washington valleys can trap smoke and gas close to the ground, nothing like a coastal site such as Moss Landing where wind disperses emissions.
Fisher Creek runs through the parcel with an unmapped floodplain less than 10 vertical feet from the development area. Contaminated firefighting runoff would reach the creek.
Snoqualmie Ridge has limited ways out: Snoqualmie Parkway and SR-18. Moss Landing evacuated 1,200–1,500 people. Snoqualmie Ridge is bigger.
Sodium-ion
Sodium-ion is a class of battery chemistries, not a single technology. Safety characteristics vary by formulation. Peer-reviewed research shows sodium-ion as a class has higher thermal runaway onset temperatures and lower pressure buildup than lithium-ion, but some sodium-ion chemistries can still experience thermal runaway under extreme conditions. The degree of improvement depends on the specific cathode, electrolyte, and cell design.
The comparison below reflects Peak Energy’s NFPP chemistry, which is what Jupiter Power has contracted for other projects.
| Technology | Fire Risk | Toxic Emissions | Noise | Grid-Scale Ready? |
|---|---|---|---|---|
| Lithium-ion (NMC) | High | High (HF, heavy metals) | High (active cooling 24/7) | Yes |
| Lithium-ion (LFP) | Moderate | Moderate (HF still produced) | High (active cooling 24/7) | Yes |
| Sodium-ion (NFPP) | Very low in abuse testing | Low (no heavy metals, much less HF) | Low (passive cooling) | Not at this scale. Peak Energy’s first utility-scale NFPP system, delivered September 2025, is a 3.5 MWh pilot. Cascadia Ridge would be 520 MWh or more |
Abuse testing
Sodium-ion cells have been tested under the same mechanical abuse conditions that cause lithium-ion cells to catch fire. The results below are encouraging, but note which company and which cathode each one comes from: CATL’s demonstrations use an NFM layered-oxide cathode and Natron’s use Prussian blue analogues, neither of which is Peak Energy’s NFPP. A bare “sodium-ion” commitment is not the same as an NFPP commitment.
- Nail penetration and mechanical abuse: CATL publicly demonstrated nail penetration, drill, and metal saw tests on sodium-ion packs without ignition. Cells also remain stable under crush testing. In comparative testing (CATARC), sodium-ion showed a 0% ignition rate vs. 23% for lithium-ion, where 85% of cells exceeded 200°C.
- UL 9540A: Natron Energy was the first sodium-ion company to publish full UL 9540A results, and its cells required no additional safety controls. UL 9540A is a test method rather than a certification: it produces a report characterizing thermal-runaway behavior and carries no pass or fail. Peak Energy has published no UL 9540A results for its NFPP cells.
Peak Energy
Peak Energy’s sodium-ion cells use an NFPP cathode (sodium iron phosphate pyrophosphate) with a non-flammable electrolyte. They’ve passed nail penetration, overcharge, and crush testing without thermal runaway. No cobalt, nickel, or manganese. Passive cooling only: no fans, no HVAC, no 24/7 noise. Peak Energy delivered its first US sodium-ion storage system in September 2025, a 3.5 MWh pilot, and is building its first domestic giga-scale manufacturing facility.
Jupiter Power announced a sodium-ion supply agreement with Peak Energy on November 12, 2025, which the parties said could be worth more than $500 million. Their CTO called it a “potential game changer.” The agreement covers a firm 720 MWh for delivery in 2027 plus a reservation of up to 4 GWh through 2030. Cascadia Ridge isn’t scheduled to come online until late 2028.
A full EIS alternatives analysis would require Jupiter Power to evaluate whether sodium-ion is feasible at this site, given the proximity to thousands of homes, a school, limited evacuation routes, and a fire district that cannot confirm preparedness for a lithium-ion fire. Without an alternatives analysis, the chemistry question never enters the public record.
Why did lithium-ion get commercialized first?
Lithium-ion and sodium-ion were both researched starting in the 1970s. Lithium-ion won the race to commercialization because of energy density. Lithium is the lightest metal on the periodic table and has the highest electrochemical potential, which means more energy in a smaller, lighter package. When Sony commercialized the first lithium-ion battery in 1991, the market was portable electronics: laptops, camcorders, eventually phones. For those applications, energy density is everything. Sodium-ion couldn’t compete on that metric and the research funding followed lithium-ion. The EV market in the 2010s reinforced the same dynamic: range anxiety means weight and size matter.
Sodium-ion got left behind because there was no market where its advantages mattered. It’s heavier and lower energy density, which is disqualifying for anything you carry or drive.
What changed is stationary grid storage. A BESS sits on the ground. Nobody cares what it weighs. The metrics that matter are cost, safety, cycle life, and supply chain stability. Sodium-ion is competitive or better on all four. Sodium is the sixth most abundant element on earth, with no cobalt, nickel, or lithium supply chain risk. The market for utility-scale stationary storage at this scale basically didn’t exist until renewable mandates like CETA created demand for it. That’s why sodium-ion is reaching commercial scale now.
Is sodium-ion unproven?
Sodium-ion does not have an operational track record at grid scale, and that gap is real. As of 2025 it is no longer hypothetical, but the deployed base is small. Peak Energy delivered the first US grid-scale sodium-ion system in September 2025. CATL began mass production of its Naxtra sodium-ion line in April 2025 and is deploying across energy storage in 2026. BYD commissioned a sodium-ion mass production line in July 2025. The technology is in production and on the grid.
The safety case for sodium-ion is also fundamentally different from lithium-ion’s.
Lithium-ion’s risk comes from thermal runaway, and you can only learn how often that happens by running facilities for years and counting incidents. That’s why the failure rate data matters and why the lack of data on aged systems is concerning.
Sodium-ion’s safety case comes from the fact that the cells don’t thermally run away in the first place. That’s testable in a lab, and it’s been tested. Peak Energy’s NFPP cells have been through the same abuse testing protocols used for lithium-ion (nail penetration, crush, overcharge, saw) with a 0% ignition rate vs. 23% for lithium-ion. Natron Energy published full UL 9540A results for sodium-ion before any commercial deployment, which is more pre-deployment safety data than lithium-ion BESS had when the industry started deploying it at scale.
The operational unknowns for sodium-ion are real: cycle degradation at scale, long-term maintenance costs, supply chain reliability. Some of that is commercial risk. But an absent operational record is also a safety-record gap, and it is why we are not asking the County to approve this project with a different chemistry. NIST’s March 2026 assessment (Technical Note 2365) found that lithium-ion fire reporting substantially underestimates real incidence, which is a reminder that a chemistry’s field record is built by running facilities and counting what happens. Sodium-ion has not yet accumulated one at this scale.
This is the second half of the problem. Lithium-ion at this site carries impacts the record has not evaluated. A first-of-its-kind NFPP deployment at 520 MWh or more would carry impacts nobody has evaluated anywhere. Neither path supports approval on the current record. That is why the ask is a full Environmental Impact Statement rather than a chemistry swap.
Further reading
- Safety of sodium-ion batteries: prospective analysis from first generation towards more advanced systems (MDPI Batteries, 2024). Finds sodium-ion has higher thermal stability than lithium-ion but notes that safety varies by chemistry and that some sodium-ion formulations can still thermally runaway.
- Thermal runaway hazards comparison between sodium-ion and lithium-ion batteries using accelerating rate calorimetry (Process Safety and Environmental Protection, 2024). Compared NTM sodium-ion, LFP lithium-ion, and NCM lithium-ion. Found sodium-ion thermal runaway characteristics closer to the safer LFP profile than to NCM.
- Natron Energy UL 9540A module test report (PDF). First sodium-ion company to publish full UL 9540A results. Cells passed without additional safety controls.