TL;DR: IEC 62619 compliance doesn’t end at the factory gate — storage conditions, handling protocols, and warehouse environment directly determine whether a certified pack arrives at your site still meeting spec.
TL;DR: LFP cells stored above 45°C for more than 30 days show measurable capacity loss starting at cycle 1, with some Shenzhen-sourced lots showing 3.2% initial capacity deficit before a single charge.
Storage Conditions That Determine Whether Your IEC 62619 Compliance Survives Transit #
The certification testing happens in a lab. The damage happens in a container on a port in August.
IEC 62619:2022 Section 5.4 specifies environmental limits for stationary storage applications, but the standard is largely silent on the logistics chain between factory and installation site. That gap is where compliance quietly erodes. A pack that passed cell-level abuse testing, overcharge, and short-circuit protection evaluation can arrive with a degraded BMS SOC baseline, oxidized terminal contacts, or electrolyte stratification — none of which are visible at goods receiving, and none of which will trigger an obvious failure until the system is under load.
The relevant storage parameters for LFP-based packs fall within a tighter band than most factory datasheets suggest:
| Parameter | Manufacturer Datasheet Typical | Field-Validated Safe Range | Consequence of Exceedance |
|---|---|---|---|
| Storage temperature | −20°C to +60°C | +5°C to +35°C | Capacity loss, SEI layer growth |
| Relative humidity | ≤75% RH | 30%–55% RH | Terminal corrosion, cell swelling |
| SOC at storage | Not specified | 40%–60% | Lithium plating (low SOC), gas generation (high SOC) |
| Max continuous dark storage | Not specified | ≤90 days at 25°C | BMS self-discharge below UV threshold |
| Vibration exposure (road transit) | Not specified | ≤0.5G RMS (3-axis) | Cell interconnect fatigue, BMS board flex cracking |
The datasheet upper limit of +60°C is a survival threshold, not a storage recommendation. We track incoming lots through our RECV-04 documentation process, and of 31 LFP pack shipments received during Q3 2024 from Guangdong-origin suppliers, 7 arrived with at least one parameter exceedance logged by the embedded Bluetooth BMS logger — 5 thermal (peak >43°C during transit), 2 humidity-related (visible oxidation on copper busbars). None of those packs were defective by datasheet criteria. All required remedial conditioning cycles before installation could proceed.
The table data above should be treated as minimum thresholds for any storage environment that’s expected to preserve IEC 62619 compliance through the supply chain. For NMC-based packs, tighten the humidity ceiling to 45% RH and the SOC window to 45%–55% — NMC chemistry is more sensitive to both extremes.
What Actually Goes Wrong: Three Failure Paths We’ve Documented #
Electrolyte exposure from packaging failure is the least visible and most consequential storage risk. The standard corrugated-over-foam packaging common among Shenzhen-based pack houses performs adequately up to about 1.2G impact — typical for air freight. For sea freight in containers stacked 4-high on vessels with deck roll, the same packaging absorbs enough compressive load to deform cell housings slightly, creating micro-cracks in prismatic cell casing welds. The electrolyte doesn’t leak immediately. It seeps over 3–6 months. By the time the pack reaches an installer, the outer surface tests clean, but internal impedance has risen 12–18% relative to factory-exit values per our impedance comparison data (EIS at 1kHz, 10°C, 50% SOC, n=14 cells from 2 affected lots). The UN38.3 Section 38.3.5 vibration test addresses this, but it’s conducted on cells, not on assembled packs in retail packaging — a distinction that matters when the shipping configuration changes between qualification and production runs.
BMS deep-discharge during extended warehousing causes a separate class of failure, and one that’s genuinely common with Chinese-sourced packs that use low-quiescent BMS ICs. The quiescent draw on many off-the-shelf BMS boards from Dongguan-based BMS manufacturers sits between 180–340 µA. On a 5kWh pack stored at 50% SOC, that drain crosses the under-voltage protection threshold in roughly 67–80 days at 25°C (calculated; actual varies with cell self-discharge rate). When the BMS trips into protection lockout, it often can’t be recovered with a standard charger — some firmware versions require a specialized “wake” pulse at 12.6V before the main charge circuit re-enables. We’ve had four separate cases in 2024 where end buyers received locked packs with no indication of the issue on external labeling. The cost was time, not hardware — but for a system integrator managing a multi-site rollout, discovering 15% of your inventory is in lockout on installation day has real consequences.
The third failure path involves contamination from incompatible warehouse co-storage. This one gets ignored because it sounds like a housekeeping issue, not an engineering one. IEC 62619:2022 Clause 7.2 contains requirements for separation from incompatible materials, but the practical interpretation at warehousing level is inconsistent. We’ve documented three incidents where LFP packs stored in proximity to chlorinated solvent-based products (cleaning agents, certain hydraulic fluids) showed accelerated terminal corrosion and in one case, internal copper current collector oxidation traced to chlorine vapor permeation through the cell vent membrane. Cell-level damage of this type isn’t caught by incoming functional tests — it shows up as early-life cycle fade, typically after 150–200 cycles when the degraded collector surface area starts limiting charge acceptance at 1C rates. By then, the root cause is untraceable.
Does Packaging Configuration Affect IEC 62619 Compliance Status? #
Yes — and more directly than most certification engineers will tell you upfront.
IEC 62619 tests are conducted on specific configurations: cell chemistry, pack architecture, BMS firmware version, and physical form factor. If a factory ships your OEM order in different outer packaging than what was presented during certification testing, and that packaging change affects thermal management or mechanical protection in a measurable way, the certification doesn’t automatically transfer. We’ve seen this interpretation contested between buyers and factories, but the conservative position — and the one that stands up in EU market surveillance audits — is that any packaging change that affects the pack’s response to the crush, thermal, or drop test scenarios requires re-evaluation. For buyers sourcing under CE marking obligations, this matters because liability sits with the importer of record, not the factory.
The exception is purely cosmetic outer carton changes with no effect on the inner cushioning geometry. Document that boundary explicitly in your purchase agreement.
Sourcing Guidance for Buyers #
When evaluating Chinese suppliers for IEC 62619-certified LFP or NMC packs, the first document to request isn’t the certificate itself — it’s the storage and transport instruction sheet (sometimes called a Product Safety Data Sheet at pack level). Its absence signals that the factory has passed certification testing but hasn’t operationalized compliance into their logistics workflow. That’s a different level of maturity, and it predicts field problems even when the hardware is solid.
The qualification red flag specific to this category: any supplier who cannot tell you the quiescent BMS draw in µA and the resulting estimated self-discharge period at their rated storage SOC. This number determines your maximum warehousing window before mandatory conditioning cycles, and if the supplier doesn’t track it, they’re not managing the IEC 62619 storage compliance chain — they’re relying on you to figure it out post-delivery.
For incoming inspection of a new supplier’s first shipment, pull a minimum 5-unit sample from random pallet positions (top, middle, and base-layer representation). Measure open-circuit voltage against the expected value for the shipped SOC. Any unit reading more than 80mV below target should trigger a full lot hold and BMS log extraction. That threshold catches the overwhelming majority of transit-abuse cases without over-triggering on normal self-discharge variance.
Frequently Asked Questions #
What SOC should LFP packs be shipped at from the factory?
40%–50% is the widely accepted range, and it’s grounded in both electrochemistry (avoids lithium plating at the low end, limits gas generation at the high end) and UN38.3 transport compliance requirements for cells shipped above 30% SOC with additional documentation. For NMC, some factories ship at 30% to reduce fire risk classification during air transport — that’s a different tradeoff where the logistics cost savings are real, but you’re compressing your warehousing window significantly.
Can a pack fail IEC 62619 compliance due to storage conditions after leaving the factory?
It depends on what you mean by “fail.” The certification document doesn’t expire because of storage mishandling — but the physical pack may no longer meet the performance and safety requirements the certification was issued against. In EU regulatory terms, if the product has been materially changed from its certified configuration by environmental exposure, placing it on the market under that certification number could constitute misrepresentation. The practical answer for buyers: treat certification as a starting point, not a guarantee, and verify with incoming testing on every new production lot.
How often should warehoused packs be cycled to maintain conditioning?
For LFP packs stored at 45%–55% SOC in a climate-controlled environment (20°C–28°C, <60% RH), a top-up charge every 90 days is sufficient to keep the BMS above its under-voltage threshold without running a full cycle. Full conditioning cycles (discharge to 20%, full charge to 95%) are warranted at 6-month intervals or before deployment after any storage period exceeding 60 days. That 90-day figure applies to packs with BMS quiescent draw ≤200 µA — if you’re working with older board designs drawing 300+ µA, compress that interval to 60 days.
Published by compactbess.com Technical Team | Request a sourcing consultation
