Canadian Solar validates 15,000-cycle sodium-ion BESS
Canadian Solar shipped 3.7 GWh of batteries in Q2 2026, up 73%, and says it is validating a sodium-ion BESS platform rated above 15,000 cycles.
Canadian Solar is validating a sodium-ion BESS platform, SolBank Na, that executive chairman and CTO Shawn Qu said will eventually deliver over 15,000 cycles. The disclosure came with Q2 2026 results published on 27 August 2026: battery shipments of 3.7 GWh, up 73% year on year, against module shipments of 3.1 GW, down 60% and the lowest since mid-2020.
Q2 2026 in numbers
| Parameter | Value | Source |
| BESS shipments, Q2 2026 | 3.7 GWh, up 73% year on year | Canadian Solar Q2 2026 results |
| Of which internal projects | 471 MWh; revenue on 3.3 GWh | Canadian Solar |
| e-STORAGE revenue / backlog | US$426m / US$3.5bn contracted | Canadian Solar |
| Projects under long-term service agreements | 34 GWh operating | Canadian Solar |
| PV module shipments | 3.1 GW, down 60% year on year | Canadian Solar |
| Net revenue | US$1.2bn, up 12% q-o-q, down 29% year on year | Canadian Solar |
| Gross margin / net result | 13.9% (vs 29.8%) / net loss US$77m | Canadian Solar |
| Q3 2026 guidance | 3.5-3.8 GW modules, 3.4-3.8 GWh BESS | Canadian Solar |
| Full-year US market guidance | 6.5-7 GW modules, 4.5-5.5 GWh BESS | Canadian Solar |
| Leadership | Colin Parkin CEO; Shawn Qu chairman and CTO | Canadian Solar |
Manufacturing produced US$1.1 billion of the quarter's revenue at an 11.9% gross margin and an operating loss of US$49 million, which CEO Colin Parkin attributed to higher unit shipping costs and ramp-up expenses. Parkin, previously head of e-STORAGE, said manufacturing "is also where our strategic priorities lie."
What SolBank Na actually is, and what it is not
SolBank Na is a containerised sodium-ion storage platform still in validation. Qu named three claimed advantages over lithium iron phosphate: abundant raw materials, simplified cooling requirements and reduced thermal runaway risk. No commercial availability date, no price point and no third-party test certificate were given. A cycle-life figure quoted on an earnings call before a product ships is a target, not a specification.
The shipping products sit alongside it. SolBank 3.0 delivers 5 MWh in a 20-foot enclosure using 314Ah LFP cells. SolBank 4.0, due in 2027, raises density 25% to 6.25 MWh in the same footprint using 588Ah cells. The company is also moving from air-cooled to liquid-cooled medium-voltage skids, packing 32 inverters of 450 kW each into 14.4 MW in a 40-foot layout, and exploring a 2.5 MW solid-state transformer stepping 34.5 kV AC to 800 V DC at over 98.5% conversion efficiency.
Why cooling and thermal runaway are the Gulf variables
Two of the three sodium-ion advantages Qu cited are exactly the two costs that scale with ambient temperature. Every battery enclosure in the Emirates runs an active cooling system, and that system consumes energy from the same asset it protects — parasitic load that reduces round-trip efficiency, worst in July and August when the system is most valuable. Fire risk drives the second cost: separation distances, detection and suppression systems, and insurance premiums.
A chemistry with a wider safe operating window and lower runaway propensity attacks both. That is why sodium-ion is worth watching from Dubai specifically rather than as a general technology story. It is also why claims about it need to be treated with more scepticism here than elsewhere: no Gulf deployment has yet demonstrated any of it in 50°C ambient.
The SOLTECH view
We think the storage-versus-modules split in these results is the more useful signal, and the sodium-ion headline is the less useful one.
- For a villa owner in the UAE: nothing here is buyable. SolBank is utility-scale containerised equipment, and SolBank Na is not a product yet. If a supplier offers you a sodium-ion home battery on the strength of a 15,000-cycle figure, ask which certificate that number comes from and which installation in the Gulf has run it. We would not specify an unvalidated chemistry on a residential system where the warranty is the entire risk transfer. Our current guidance for villa systems stays with LFP.
- For a commercial or industrial owner: the number to note is 588Ah. Higher-capacity cells mean more stored energy per square metre of plant room or yard, and space is a real constraint on C&I sites in the UAE. If your storage project has a 2027 or later commissioning date, it is worth asking bidders which cell generation they are quoting, because a 25% density gain arriving mid-procurement changes the layout.
- For an EPC contractor and investor: a second major manufacturer has now reported storage growing while modules collapse — Canadian Solar's batteries up 73% and modules down 60%, after Sungrow's storage revenue passed its inverter revenue. That is a pattern, not a coincidence, and it means storage supply chains are where the vendor competition and the pricing pressure will be over the next two years. It also means module supply is consolidating into fewer, larger and less discount-prone hands.
Where we are cautious: a 13.9% gross margin and a US$77 million net loss are not the financials of a company in a position to fund a new chemistry to commercial maturity on a short timeline. We would place SolBank Na's realistic Gulf availability well beyond 2027 and would not let it into any procurement schedule until a certificate and a reference site exist. See the equipment we currently specify.
Sources: Energy-Storage.news — Canadian Solar ships 3.7GWh of BESS in Q2 2026, 2 September 2026, PV Tech — Canadian Solar Q2 PV module shipments fall 60% year-on-year, Canadian Solar investor relations.