Technical Evaluation & Sample Request Guide for Cycle Life & Degradation

What to Specify Before You Send a Single Inquiry #

The quality of data you get back from a Chinese cell supplier is almost entirely a function of how you ask. Generic sample requests (“please send your best LFP 280Ah cell for evaluation”) produce generic responses: a marketing datasheet, a COA with round numbers, and a sales contact who will not answer technical questions directly.

Before drafting your inquiry, lock down these five parameters internally:

• Discharge rate for cycle life: Specify 0.5C/0.5C or 1C/1C — not just “standard cycling.” Suppliers default to 1/3C unless told otherwise, and the cycle life delta is material. A cell rated at 4,000 cycles at 1/3C may deliver 2,847 cycles at 1C/1C under the same 80% EOL threshold.
• Temperature range: State your operating range explicitly (e.g., 15°C to 45°C ambient). Cycle life data at 25°C is a baseline, not a deployment condition.
• State-of-charge window: If your application cycles between 20% and 90% SOC rather than 0–100%, say so. Partial SOC cycling changes degradation mechanics and some suppliers have this data; most won’t volunteer it unless asked.
• End-of-life threshold: 80% capacity retention is the IEC 62133-2 industry default, but some applications tolerate 75% or require 85%. State yours upfront.
• Required certifications and their scope: Don’t just ask “do you have UN38.3?” Ask whether the report covers the exact cell model, form factor, and nominal capacity you’re sampling. Certificate scope mismatches are one of the more common delays we see in the pre-qualification stage.

This framing forces suppliers to either confirm they can match your conditions or surface the gap early. Either outcome is useful.

Head-to-Head Comparison — Cycle Life Data Quality Across Supplier Response Types #

When we receive evaluation packages from Shenzhen-based pack houses and cell manufacturers, we score the technical response quality before we even open the physical sample box. Over 31 supplier evaluations logged under our internal QC-07 Sample Intake process between 2023 and 2024, four distinct response types emerged:

The pattern is consistent: the suppliers who can describe their test methodology in the initial response are the same ones whose lab data correlates with your own incoming test results. Trading companies and small assemblers almost never have cycle test infrastructure in-house — they’re copying numbers from the cell OEM’s datasheet, which may not apply to the specific grade or lot they’re actually shipping.

For the most common use case in our evaluations — portable power station integration requiring 2,000+ cycle life with a 12-month warranty buffer — the Tier-1 cell OEM path is the right one, even if lead time and MOQ are less flexible. A mid-tier integrator can sometimes bridge the gap if you can get access to their upstream cell supplier’s test reports and verify lot alignment.

That said, for lower-stakes applications (demonstration units, single-geography pilots under 500 units), a mid-tier integrator with a strong COA track record is often acceptable. The calculus shifts when you’re committing to production volumes above 2,000 packs.

The Overlooked Variable — Lot-to-Lot Consistency, Not Average Performance #

Cycle life datasheet values are averages. What actually determines your warranty exposure is the variance across production lots.

We’ve seen this play out concretely: a supplier’s 2023 evaluation samples tested at 3,412 cycles to 80% EOL under 0.5C/0.5C at 25°C. Strong numbers, passed qualification. Six months into production supply, warranty returns started appearing from units assembled with Q1 2024 lots. Our investigation traced the degradation to a cathode coating process change the supplier had not disclosed. The affected lots tested at 2,190 cycles — a 36% reduction from the qualified baseline.

This is why lot-traceable cycle data matters more than headline performance. When requesting samples, specify that you need cells from a minimum of 3 separate production lots, not just 3 individual cells from the same lot. For battery pack design decisions involving multi-cell configurations, inter-lot consistency in initial impedance is equally important: a spread of more than 2 mΩ across a single-lot sample of 18 cells is a signal worth escalating before you commit to pack assembly.

IEC 62619 Section 5.4 addresses manufacturing consistency requirements for secondary lithium cells in industrial applications, but it sets a compliance floor, not a performance ceiling. Suppliers who cite IEC 62619 compliance as evidence of lot consistency are conflating two different things.

Ask specifically: “Can you provide cycle life test data from three non-consecutive production lots within the past 12 months?” If they can’t, that absence tells you something about their internal QA infrastructure, not just their data management.

Implementation Notes — What to Watch for After You Receive Samples #

Once samples arrive, the incoming inspection sequence matters as much as the cycle test itself. We run a staged protocol that prioritizes quick-turnaround checks before committing cells to the long-duration cycle test:

Day 1–2 (Physical and electrical baseline):
– Dimensional verification against datasheet (thickness variation >0.3mm on prismatic cells warrants a hold)
– Open circuit voltage check: all cells should be within ±15mV of each other out of the box
– DC internal resistance (DCIR) at 50% SOC using a 10-second pulse at 0.5C, per IEEE 1679.1 guidance on characterization methods for secondary lithium-ion cells

Week 1–2 (Formation cycling and capacity verification):
– Three formation cycles at 0.2C charge/discharge before recording reference capacity
– Record capacity at C/5, C/2, and 1C — the ratio between C/5 and 1C capacity tells you a lot about the cell’s internal resistance profile. A ratio below 0.91 on an LFP cell is worth flagging.

Weeks 3–14 (Accelerated cycle test):
– Run at your target C-rate, your target temperature, your SOC window
– Check at cycle 100, 300, 500 for early-trend extrapolation
– Any cell showing more than 4% capacity fade by cycle 100 under standard conditions should be flagged immediately — don’t wait for the 500-cycle checkpoint

Common early-shipment red flags: inconsistent terminal tab welding visible under inspection, electrolyte odor at the seam (indicates a sealing defect), and COA capacity values that are suspiciously uniform across all cells (round numbers like exactly 280.0Ah on every cell usually mean the figure wasn’t measured on that specific lot).

Build in a hard milestone: if you don’t have cycle trend data through at least 300 cycles before your design-in freeze deadline, push the freeze. A 2-week extension to your evaluation timeline is far cheaper than a field recall.

For applications where BMS engineering parameters will be tuned to the specific cell chemistry, the impedance and capacity ratio data from this incoming inspection feeds directly into SOC algorithm calibration — don’t treat it as a standalone QC step.

Sourcing Guidance for Buyers #

When evaluating Chinese suppliers in this category, the first document to request is the cycle test report with its raw data file, not just the summary page. A legitimate in-house test will include: equipment ID, temperature log, per-cycle capacity values, and the cell serial numbers tested. Suppliers who can only provide a one-page summary table — with no equipment reference and no per-cycle data — are almost certainly sharing a manufacturer’s reference report, not results from their own testing. That distinction is significant when the cells you receive may be a different grade or lot than the ones tested.

The qualification red flag specific to cycle life claims: a supplier who quotes a single cycle life number without any stated EOL threshold. “Our cells last 4,000 cycles” is not a specification. Cycle life without an EOL definition (80% capacity retention is standard under IEC 62133-2) is unverifiable and unenforceable in any supply agreement.

For incoming inspection, our standard is to pull 18 cells from each incoming lot and run capacity verification at 0.2C. Accept the lot if all 18 cells are within ±1.8% of the stated nominal capacity. Flag for hold if more than 2 cells fall outside that band. Reject if any cell is more than 3.5% below nominal. These thresholds aren’t universal — for high-reliability applications, tighten the band to ±1.2%.

---

Published by compactbess.com Technical Team | Request a sourcing consultation