TL;DR: When requesting energy density / power density evaluation samples from Chinese cell and pack suppliers, the parameters you specify in your inquiry determine 80% of whether the data you receive is actually useful for a design-in decision.
TL;DR: In our incoming evaluation process, we require a minimum 5-sample lot for capacity verification and reject any supplier whose DCIR spread across that lot exceeds 8% at the same SOC reference point.
What to Specify Before You Ever Send the Inquiry #
Most evaluation sample requests fail before the samples ship — because the inquiry itself is underspecified. A vague “please send us your 280Ah LFP cell samples with datasheet” will get you a box of cells and a glossy PDF that tells you almost nothing useful.
The inquiry document should read like a mini-specification. At minimum, lock down the following parameters before contacting any Shenzhen-based pack house or Dongguan cell distributor: nominal voltage window (e.g., 2.50V–3.65V for LFP), target gravimetric energy density in Wh/kg and volumetric energy density in Wh/L at the system level, required continuous discharge rate (C-rate), peak pulse discharge rate with duration, and operating temperature range. If your application involves anything below 0°C, state it explicitly — most datasheets quote room-temperature figures and cold-temperature performance is rarely volunteered.
For power density-sensitive applications like power tools or high-rate portable systems, also specify minimum acceptable DCIR (DC internal resistance) at 50% SOC, 25°C. We’ve seen inquiry responses where suppliers quote DCIR at 100% SOC, which is systematically lower and entirely misleading for real operating conditions.
One internal document we run every sample request against is our EVL-03 inquiry checklist — it covers 22 technical fields that must be confirmed before samples are dispatched. If a supplier can’t respond to more than half the fields, that alone tells you something about their engineering depth.
Interpreting the Sample Data You Actually Receive #
| Parameter | What to Measure | Acceptance Threshold | Reject If |
|---|---|---|---|
| Initial capacity (0.2C, 25°C) | 3 charge/discharge cycles, average | ≥98% of rated capacity | <95% on any cell |
| DCIR spread (50% SOC, 25°C) | 1kHz AC impedance, 5-cell lot | Cell-to-cell spread ≤8% | Spread >8% or no method stated |
| Gravimetric energy density | Weigh cell, compute from measured capacity | Within 3% of datasheet claim | >5% deviation from spec |
| Cycle retention at 1C/1C | 100-cycle interim check | ≥97% at cycle 100 | <95% at cycle 100 |
| Self-discharge | Rest 7 days at 50% SOC, 25°C | Voltage drop ≤15mV | >25mV drop |
The table above reflects the thresholds in our standard incoming cell evaluation protocol, based on assessment of 23 incoming lots over the past 18 months.
What the data will tell you: a cell can look excellent at 0.2C and fall apart at 1C discharge. Gravimetric energy density figures from Chinese suppliers are frequently quoted at 1/3C or even 0.1C conditions — which inflates the number by 4 to 9% compared to your real application rate. Verify the C-rate used for capacity measurement in every datasheet, every time. If the footnote is absent, ask for it. Blank footnotes on capacity test conditions are not an oversight.
For the cycle retention check, 100 cycles is not a complete qualification — it’s a screening gate. A cell that drops below 97% retention at cycle 100 under IEC 61960-3 test conditions is almost certainly not going to hit the 2,000-cycle retention figure claimed on the front page of the datasheet. You don’t need to run 2,000 cycles before making a preliminary go/no-go call.
The UN38.3 test report is worth requesting at this stage too — not because you need it for design-in, but because how a supplier handles that request tells you whether their quality infrastructure is real or performative.
What Goes Wrong Between Datasheet and Delivered Sample #
This is where most evaluation programs stall or produce misleading conclusions, and the failure modes are predictable.
The first scenario involves energy density inflation through cell grading mismatch. A supplier quotes 270 Wh/kg gravimetric energy density based on their Grade-A production cells. What arrives in your evaluation sample box are cells from a secondary grading bin — capacity is 3.2% lower, internal resistance is 11% higher, and the weight is within tolerance. The datasheet numbers are technically “real” but apply to a different grade than what was sent. When you later scale to production quantities, the Grade-A cells are allocated to a higher-tier customer and your production supply defaults to the same secondary bin. We’ve seen this pattern with pack suppliers in Huizhou, particularly for branded-cell orders where the upstream cell source isn’t contractually locked. The check: request a cell traceability lot code that links to the original cell manufacturer’s outgoing QC record, and verify it before design-in approval.
The second failure mode is power density misrepresentation through pulse rating ambiguity. A supplier’s datasheet states peak discharge: 3C. What they mean is 3C for 5 seconds at 25°C, starting from 80% SOC. Your application requires 3C for 30 seconds, starting from 30% SOC, at 10°C. Those are completely different thermal and electrochemical conditions. The cell that passed their own 3C spec will voltage-sag below your cutoff threshold in under 15 seconds at low SOC and low temperature. The consequence in a portable power station context is that your BMS triggers an under-voltage protection fault under peak load — which your end customer experiences as a sudden shutdown. The IEC 62619:2022 standard, Section 5.4 addresses operating condition disclosure requirements, but compliance there doesn’t mean a supplier will proactively clarify their pulse test conditions. You have to ask, and you have to specify your actual use condition in the inquiry.
The third scenario is one that rarely gets discussed in supplier qualification guides. A buyer working with a Shenzhen cell distributor — not a direct cell manufacturer — requested samples for a high-power application targeting 450 W/kg power density. The distributor provided samples with a correctly measured DCIR of 0.28 mΩ per cell at 50% SOC, 25°C. Passed intake. Approved. Production order placed. Three months later, the production batch DCIR measured at 0.41 mΩ — a 46% increase — because the distributor had switched upstream cell source between sample supply and production supply without disclosure. The buyer had no contractual cell-source lock in the supply agreement. Total redesign cost and delayed launch added up to roughly $94,000 in direct costs. The DCIR threshold spec was right. The contract clause protecting it was missing. Under IEEE 1725 guidelines for cell qualification in portable applications, cell-source substitution should be treated as a material change requiring buyer notification — but that obligation needs to be written into your purchase agreement explicitly.
Does Energy Density or Power Density Matter More for Your Application? #
It depends on discharge duration. For applications discharging over more than 2 hours (0.5C or lower), gravimetric energy density is the design constraint. For anything discharging in under 30 minutes at peak, DCIR and thermal performance under pulse load matter far more than rated Wh/kg.
The practical boundary sits around 1C continuous. Below that, optimize for energy density. Above that, evaluate the power density figure at your actual operating temperature and SOC window — not the room-temperature, full-charge peak value. For battery pack design decisions in portable systems, this distinction drives cell selection more than any single datasheet headline number.
Sourcing Guidance for Buyers #
When evaluating Chinese suppliers in this category, the first document to request is the cell manufacturer’s outgoing QC report — not the distributor’s own in-house test report. The absence of an original cell-level QC certificate, with lot codes traceable to the cell manufacturer’s own production records, is a signal that the distributor has limited transparency into their upstream supply chain. That’s not always disqualifying, but it means your sample data may not represent production supply.
One qualification red flag specific to this category: any supplier who provides energy density and power density figures without specifying the C-rate, temperature, and SOC conditions under which they were measured. A 280Ah LFP cell that shows 163 Wh/kg at 0.1C and 25°C may show 149 Wh/kg at 1C and 10°C. Both numbers are real. Only one matches your application.
For incoming inspection, our standard protocol is a 5-cell minimum lot with 100% DCIR measurement at two SOC reference points (50% and 80%), plus a 3-cycle capacity run on at least 3 of the 5 cells. If DCIR spread across the 5 cells exceeds 8% at either reference point, the lot goes to secondary review under our IQC-11 hold procedure before any design-in decision proceeds.
For the broader context on how BMS engineering interacts with cell-level energy and power density specs — particularly SOC window management and discharge cutoff thresholds — that’s a parallel qualification track that should run alongside cell evaluation, not after it.
As of mid-2025, Grade-A LFP prismatic 280Ah cells from credible Shenzhen distributors with traceable cell-manufacturer QC documentation trade at $0.056–0.063 per Wh ex-works. Evaluation sample pricing is typically 15–25% above that. If you’re seeing ex-works pricing below $0.049/Wh on “Grade-A” cells, ask for the lot traceability documentation before proceeding.
Timeline from a well-specified inquiry to a design-in decision typically runs 6–9 weeks: 1 week for supplier response and sample dispatch confirmation, 2 weeks for shipping and customs clearance, 2 weeks for intake testing, and 1–3 weeks for cycle screening and data review. Any supplier who tells you they can turn a credible evaluation around in under 4 weeks is either skipping cycle testing or sending you pre-certified samples that aren’t from your actual production lot.
Frequently Asked Questions #
How many samples do I need for a valid energy density evaluation?
Five cells minimum for statistical DCIR spread analysis; if you’re running full cycle testing, pull 3 cells for destructive cycling and retain 2 as reference samples sealed at 50% SOC for later comparison.
Should I test at my application C-rate or the datasheet C-rate?
Your application C-rate, always. Datasheet conditions are a supplier’s best-case scenario, not your operating scenario. If you only have time to run one capacity test, do it at 1C discharge, 25°C — it’s a reasonable middle ground that exposes rate-dependent capacity loss without requiring a specialized fixture.
Can I trust the cycle life numbers on the datasheet?
You can treat them as a ceiling, not a floor. Cycle life claims on Chinese cell datasheets are almost universally measured at 0.5C/0.5C, 25°C, with a 100% depth of discharge. If your application cycles at 1C/1C, operates at 35°C ambient, or uses partial SOC windows, your realized cycle life will differ — sometimes significantly. The IEC 61960-3 standard defines a reference cycle life test methodology; check whether the supplier’s datasheet references it or uses an internal method.
What if the supplier refuses to share cell-manufacturer QC lot data?
It depends on the supplier tier. Tier-1 cell manufacturers (CATL, EVE, CALB) won’t share raw lot data directly with distributors’ end customers — that’s normal. What you should be able to get is a distributor-issued certificate of conformity that maps their sample lot code to the cell manufacturer’s outgoing test code. If a supplier can’t produce even that, the traceability chain is broken.
Is it reasonable to ask for impedance spectroscopy data, not just 1kHz DCIR?
For most procurement evaluations, 1kHz DCIR is sufficient. Full EIS (electrochemical impedance spectroscopy) data becomes relevant if you’re qualifying cells for extreme temperature applications or trying to distinguish calendar aging effects — that’s more of a second-stage qualification activity than a first-pass screen.
How do I handle a supplier who sends better samples than what they can supply at volume?
Build a contractual cell-source lock into your purchase agreement that specifies the cell manufacturer name, model designation, and grade classification. Then run a requalification DCIR check on the first three production lots using the same 5-cell protocol from your initial evaluation. A spread increase of more than 3% relative to baseline is grounds for a supplier corrective action request before you release payment.
What’s the right point to transition from evaluation to production supply agreement?
After you have passing results from at least 100 charge/discharge cycles at your target C-rate, confirmed lot traceability, and a documented DCIR baseline. Skipping the cycle screen to accelerate time-to-market is a legitimate business decision — but frame it explicitly as a risk acceptance, not a qualified evaluation. The distinction matters when something fails in the field.
Published by compactbess.com Technical Team | Request a sourcing consultation
