{"id":284979,"date":"2026-08-16T17:54:45","date_gmt":"2026-08-16T15:54:45","guid":{"rendered":"https:\/\/www.heiwit.com\/?p=284979"},"modified":"2026-08-16T17:55:38","modified_gmt":"2026-08-16T15:55:38","slug":"polyanionic-nfpp-cells","status":"publish","type":"post","link":"https:\/\/www.heiwit.com\/en\/celle-polianioniche-nfpp\/","title":{"rendered":"NFPP polyanionic cells: the second generation of Heiwit sodium-ion batteries"},"content":{"rendered":"

Sodium-ion batteries are not a single technology. The same name encompasses very different cathode chemistries, with varying characteristics in terms of lifespan, safety and performance in cold conditions. The first generation of Heiwit systems uses a cathode made of layered transition metal oxides<\/strong> (N\/Ax<\/sub>TMO2<\/sub>The second generation, which goes on sale from November 2026, adopts a cathode polyanionic NFPP<\/strong>.<\/p>\n

This article explains what changes on an electrochemical level and what can be gained from a stationary storage system, complete with qualification data for the new cells.<\/p>\n

What does \u201cpolyanionic\u201d mean\u201d<\/h2>\n

In a layered oxide cathode, sodium ions move between transition metal oxide planes stacked on top of each other. It is an efficient structure \u2014 it packs a lot of energy \u2014 but the planes expand and contract with each charge and discharge cycle, and over time they lose order. This is the mechanism that limits the service life of this family of cells.<\/p>\n

In a polyanionic cathode, sodium instead moves inside a rigid three-dimensional lattice<\/strong>, held together by phosphate and pyrophosphate groups bound to iron. The abbreviation NFPP stands for sodium iron pyrophosphate phosphate<\/em>, Na4<\/sub>Faith3<\/sub>(PO4<\/sub>)2<\/sub>(P2<\/sub>O7<\/sub>): a scaffold that deforms very little when sodium goes in and out.<\/p>\n

The consequence is a precise compromise, and it is worth stating it clearly because it is the heart of the choice:<\/p>\n

\n

The polyanionic framework stores less energy per kilogram<\/strong> of a layered oxide, but returns it for many more cycles<\/strong>, with greater thermal stability and better performance at extreme temperatures.<\/p>\n<\/blockquote>\n

It is a general property of the polyanionic cathode family, not a single manufacturer's data: lattice rigidity costs mass, and that mass does not take part in energy storage.<\/p>\n

For an electric car it would be an unfavourable trade-off: weight matters there. For a stationary storage system leaned against a wall or in a technical room, where no one lifts the battery and you still want it working in twenty years\u2019 time, it is the right trade-off.<\/p>\n

The rest of the cell remains unchanged<\/h2>\n

It is worth clarifying what not<\/em> change, because it is the reason why the process is industrially manageable. The anode remains in hard carbon<\/strong>, hard non-graphitising carbon, the same as the first generation. The operating principle is identical. The manufacturing process is too: cathode, anode and electrolyte are swapped on the same line, without redesigning the plant.<\/p>\n

One particular feature that sodium carries over from both chemistries deserves a note: unlike lithium, which requires a copper current collector at the anode, in sodium cells both manifolds can be made of aluminium<\/strong>. It is not just a question of the cost of the raw material. It is what makes it possible to discharge the cell completely down to zero volts in order to store and transport it safely, without damaging it \u2014 something that a lithium cell cannot tolerate.<\/p>\n

The figures, side by side<\/h2>\n

The table compares NFPP cells with the benchmark chemical types in the energy storage market. The values for lead-acid, LFP and NMC\/NCA are typical ranges; the NFPP values are the qualification data declared for the cells used by Heiwit.<\/p>\n

\n\n\n\n\n\n\n\n\n\n\n\n
Parameter<\/th>\nLead-acid<\/th>\nLFP<\/th>\nNMC\/NCA<\/th>\nNFPP
(polyanionic sodium)<\/th>\n<\/tr>\n<\/thead>\n
Energy density (Wh\/kg)<\/td>\n30\u201350<\/td>\n120\u2013180<\/td>\n220\u2013280<\/td>\n100\u2013120<\/strong><\/td>\n<\/tr>\n
Life cycles<\/td>\n300\u2013500<\/td>\n> 6,000<\/td>\n800<\/td>\n9.000<\/strong> (DoD 95%)<\/td>\n<\/tr>\n
Residual capacity at \u221220 \u00b0C<\/td>\n< 60%<\/td>\n< 70%<\/td>\n> 70%<\/td>\n> 90%<\/strong><\/td>\n<\/tr>\n
Residual capacity at \u221240 \u00b0C<\/td>\n0%<\/td>\n0%<\/td>\n0%<\/td>\n> 85%<\/strong><\/td>\n<\/tr>\n
Energy efficiency<\/td>\n-<\/td>\n95%<\/td>\n-<\/td>\n97%<\/strong><\/td>\n<\/tr>\n
Deep discharge tolerance<\/td>\npoor<\/td>\npoor<\/td>\npoor<\/td>\nGreat<\/strong><\/td>\n<\/tr>\n
High-temperature behaviour<\/td>\nhalf<\/td>\ngood<\/td>\nhalf<\/td>\nGreat<\/strong><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n

Two lines deserve careful reading.<\/p>\n

La energy density<\/strong> It sits below that of lithium iron phosphate: 100\u2013120 Wh\/kg. For the same installed capacity, a polyanionic sodium system is bulkier and heavier than an LFP one. For fixed domestic or commercial storage, this translates into a few centimetres and a few kilos extra, not a usage limitation \u2014 but it is right to know beforehand, not afterwards.<\/p>\n

I cycles<\/strong> are declared in 9,000 at the discharge depth of the 95%<\/strong>. This condition must be considered alongside the figure, as it is what makes the data comparable: many technical data sheets state the number of cycles at lower depths of discharge, where the figure naturally increases. Nine thousand cycles at a DoD of 95% means that the cell is designed to be discharged almost completely every day, which is exactly how a storage system paired with a photovoltaic installation works: it charges during the day and discharges in the evening. At one cycle per day, that equates to more than twenty years of service.<\/p>\n

Cold weather behaviour<\/h2>\n

It is the most visible difference compared to lithium iron phosphate, and the easiest to verify in the field.<\/p>\n

In the descending temperature discharge test \u2014 charging at 3.5 V with 0.5P power, 10 minutes rest, then discharging at the test temperature \u2014 the NFPP cell retains the 95.041 TP16T capacity at \u221220 \u00b0C<\/strong>. A reference LFP cell, in the same test, stops at 67,8%<\/strong>.<\/p>\n

In practical terms: on a January morning at \u221210 \u00b0C in an unheated plant room, an LFP battery delivers a significantly reduced fraction of its rated capacity, precisely on the days when solar panels produce less and the storage system is needed most. The gap widens as the temperature drops: at \u221240 \u00b0C, the polyanionic chemistry remains above 85%, whilst LFP and NMC have no usable capacity.<\/p>\n

Current, power and heat<\/h2>\n

The rising current discharge test \u2014 charging at 0.5C, rest, then discharging at 1C, 2C and 4C, voltage window 1.5\u20133.5 V \u2014 highlights the second structural advantage.<\/p>\n

\n\n\n\n\n\n\n\n\n
discharge regime<\/th>\nDelivered capacity<\/th>\nMaximum overtemperature<\/th>\n<\/tr>\n
NFPP<\/th>\nLFP<\/th>\nNFPP<\/th>\nLFP<\/th>\n<\/tr>\n<\/thead>\n
1C<\/td>\n100%<\/td>\n100%<\/td>\n3.5 \u00b0C<\/td>\n4.2 \u00b0C<\/td>\n<\/tr>\n
2C<\/td>\n99,4%<\/td>\n95,9%<\/td>\n5.8 \u00b0C<\/td>\n9.7 \u00b0C<\/td>\n<\/tr>\n
4C<\/td>\n98,2%<\/strong><\/td>\n82,8%<\/td>\n15.5 \u00b0C<\/strong><\/td>\n22.1 \u00b0C<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n

At 4\u00b0C, the polyanionic cell delivers almost all of its capacity and heats up by 15.5 \u00b0C; the LFP cell delivers 83% of its capacity and heats up by 22.1 \u00b0C. Less heat generated for the same current means a system that is better able to protect itself, even before the BMS or ventilation kicks in.<\/p>\n

The measured round-trip energy efficiency is 97%<\/strong> at a 0.25P rate. On a domestic storage system that cycles every day, every single point of efficiency is energy that is not lost as heat during every single cycle, for the entire lifespan of the system.<\/p>\n

The safety tests<\/h2>\n

The cells have been qualified in accordance with the test standards for stationary energy storage systems. We report the tests and the outcome as shown in the reports.<\/p>\n

\n\n\n\n\n\n\n\n\n\n\n
Test<\/th>\nReference standard<\/th>\nCriterion<\/th>\nResult<\/th>\n<\/tr>\n<\/thead>\n
Nail penetration<\/td>\nT\/CIAPS0031-2023<\/td>\nno fire, no explosion<\/td>\nsurpassed<\/td>\n<\/tr>\n
Overload<\/td>\nGB\/T 44265-2024<\/td>\nno fire, explosion or rupture<\/td>\nsurpassed<\/td>\n<\/tr>\n
Short circuit<\/td>\nGB\/T 44265-2024<\/td>\nno change in appearance<\/td>\nsurpassed<\/td>\n<\/tr>\n
Squash<\/td>\nGB\/T 44265-2024<\/td>\nno fluid leak, no fire<\/td>\nsurpassed<\/td>\n<\/tr>\n
Thermal control<\/td>\nGB\/T 44265-2024<\/td>\nno fire, explosion or rupture<\/td>\nsurpassed<\/td>\n<\/tr>\n
Bake test<\/td>\nGB\/T 36276-2018<\/td>\nno fluid leak, no fire<\/td>\nsurpassed<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n

It is worth describing two tests in full, because \u201cpassing the nail test\u201d without the conditions means nothing.<\/p>\n

Nella nail penetration<\/strong> the cell is at 25 \u00b0C and fully charged. A 5 mm diameter steel needle passes through it at the geometric centre at 25 \u00b1 5 mm\/s and it stay inside<\/em>, short-circuiting the layers for one hour of observation. The requirement is no fire and no explosion.<\/p>\n

Nell thermal control<\/strong> The charged cell is heated whilst being charged at a constant current of 1C, until it goes into thermal runaway, or until it reaches 300 \u00b0C, or until four hours have elapsed. Here too, the requirement is no fire, no explosion, no rupture.<\/p>\n

In terms of certification, the cells have reports IEC 62619<\/strong> (secondary cell safety for industrial applications), UL 9540A<\/strong> (thermal runaway propagation in energy storage systems), UN 38.3<\/strong> (transport suitability), RoHS and material safety data sheets, as well as certifications for sea and air transport.<\/p>\n

Cell construction<\/h2>\n

Prismatic cells adopt a process of stacking<\/strong> (stacking) instead of winding: the electrodes are stacked in flat layers rather than being rolled. This results in better space utilisation, lower internal impedance and less heat generation for the same current.<\/p>\n

Cylindrical cells instead adopt a design tabless<\/strong>, without connection tabs: the entire edge of the electrode acts as a collector. The effect is a very low internal resistance and distributed thermal dissipation, so much so that this construction withstands discharge pulses of up to 50C.<\/p>\n

What changes for those who already have a Heiwit battery?<\/h2>\n

Nothing changes. The layered oxide systems currently on sale and already installed retain the performance and warranty they came with; they remain supported and continue to operate with the same inverter, the same app and the same BMS. They are a mature and field-proven technology, and the cells they incorporate meet the specifications stated in their respective technical data sheets.<\/p>\n

NFPP cells will enter new production systems from November 2026, once current stock runs out. Anyone buying today is not buying an outgoing product: they are buying the generation in service, with its warranty and installation history.<\/p>\n

Insights<\/h2>\n

Three aspects of this chemistry deserve separate discussion, along with their respective protocols and test data:<\/p>\n