How to Spot Rebuilt B-Grade Cells in OEM Power Bank Quotes

1. Where B-grade and rebuilt cells come from

What cell grading actually means

Every large cell manufacturer runs a grading line at the end of production. Cells that fail the first screen, on capacity deviation, internal resistance, or self-discharge, are downgraded to B-grade and sold at a discount into secondary markets. Nothing about a B-grade cell is hidden: wholesalers list them openly as "Grade B LiFePO4 / Li-ion cells". The fraud happens one step later, when a pack assembler buys them, re-wraps them, and quotes them to you as Grade A (per industry grading conventions, not a standardized rating system).

The rebuilt-cell economy

A rebuilt cell is a used cell re-wrapped to look new, fresh film hides wear, a reprinted date code hides age.: it is a cell with history, pulled from rejected packs, recovered returns, or production overflow, cleaned, re-wrapped in fresh shrink film, and re-labelled with a new date code. Fresh film hides wear; a reprinted date code hides age. On the quote sheet both look identical, which is why quote-stage verification has to rest on documents and measurements, not on photographs.

Signal	Grade A (first-run)	B-grade / rebuilt
Quote level (10,000mAh, 500 pcs)	$5.80-8.00 FOB	15-25% below Grade-A
Cell documentation	Brand, model, datasheet on request	"High quality", no model number
Batch traceability	Date codes match packaging	Reprinted codes, mixed batches
Field return rate	<0.5%	5-8%

2. The 7 cell-level checks before you pay

Run all seven at quote stage, they take under two hours

These checks are ordered by cost: the first four cost nothing but an email, the last three need samples. Together they expose the two fraud patterns, downgraded cells sold as Grade A, and used cells sold as new.

#	Check	What a passing answer looks like
1	Cell brand + exact model	Named brand and model (e.g. LG, Samsung SDI, Panasonic), matching the quote line item
2	Datasheet	Original datasheet with capacity, IR, and cycle-life curves for that model
3	Packaging photos with date codes	Cell trays photographed with intact, uniform date codes, not re-wrapped film
4	Price sanity check	Quote within the Grade-A band; >25% below market means the cells are not first-run
5	Sample discharge test	Delivered mAh within datasheet tolerance at rated current (see section 3)
6	Internal resistance report	IR values per cell within the datasheet band, measured on a batch sample
7	UN38.3 report cross-check	Report names the exact cell model and matches the cells in your samples

Why checks 1-4 filter most fraud before sampling

A rebuilt-cell operation scales through vagueness: no model number on the quote, no datasheet, "high-quality cells" in place of a brand. Requiring all three in writing costs the supplier minutes and costs you nothing, and the suppliers who stall on checks 1-3 will not pass checks 5-7 either. Treat the price sanity check as the arithmetic backstop: cells are 35-50% of the BOM, so a quote far below the Grade-A band has to be cutting somewhere.

3. The discharge test that catches re-celled batches

Measure delivered capacity, not claimed capacity

Delivered capacity at the 5V output, not the rated cell figure, is what your buyer experiences. Delivered mAh is always lower than the cell figure because of conversion losses; that gap is normal. What exposes rebuilt cells is the pattern across units and across cycles: Grade-A cells cluster tightly within datasheet tolerance; re-wrapped used cells scatter widely and fade from the first repeated cycle.

The four-step protocol

Five units, three cycles, one datasheet: the discharge protocol takes an afternoon and turns cell grade from a claim into a measurement.

1. Sample five units from the supplier's stock batch, not from a hand-picked demo box.
2. Charge and discharge at rated current through the 5V output, logging delivered mAh per unit.
3. Repeat for three cycles, rebuilt cells lose measurable capacity between cycles; first-run cells do not.
4. Compare against the datasheet, a spread wider than the datasheet tolerance band, or visible fade by cycle three, ends the conversation.

In the case that opened this article, the delivered figure came in at 17,100mAh against a 20,000mAh rating, 85.5% delivered before conversion losses were even considered, with visible fade by cycle three. That is the rebuilt-cell signature: not one bad number, but a bad pattern.

4. UN38.3 and documentation cross-checks

One report, one cell type, match all three layers

UN38.3 is the transport safety test every lithium cell must pass before air or sea shipment, and a report costs $1,000-2,500 for a specific cell type. That specificity is what makes it a verification tool: the report names the exact cell model tested. Cross-check three layers, the cell model on the report, the cells physically inside your samples (open one unit if the volume justifies it), and the batch date codes. All three must name the same cell.

The mismatch patterns that matter

Two patterns recur in the field. The first is a borrowed report: the factory shows a genuine UN38.3 document for a well-known cell while the pack contains something cheaper, the paperwork describes a cell you are not buying. The second is a report for an unidentifiable lab, which cannot be traced if customs or your marketplace platform asks. Both fail the same way under scrutiny: ask which laboratory issued the report and which cell model it covers, then compare with what ships. For the transport-side rules your shipment will face next, our guide to dangerous goods packaging certification walks through the full compliance chain.

5. What B-grade actually costs you

The return-rate arithmetic

For a 2,000-unit batch of 10,000mAh units, the B-grade saving on the quote is roughly $2,400 (15-25% off a $5.80-8.00 Grade-A band). Processing a 5-8% return rate means handling 100-160 defective units: return freight, triage, refunds or replacements, and the review damage on a marketplace listing. At realistic handling costs the return bill alone is several times the quote saving, before counting the warranty obligation and the listing's conversion drop.

A trajectory built on Grade-A cells

From 500 units to 10,000+ annually in two years, at a 4.5-star rating across 900+ reviews. The total entry cost was $8,000-18,000, product, branding, certification, and the first batch included. The scaling worked because the return rate stayed near 0.5% instead of eating the margin: cell grade is not a detail of the BOM, it is the assumption the whole business model rests on.