How to Evaluate Roofing Slate Quality: The 6 Checks Most Buyers Skip

By the inspection team at a Jiujiang slate factory — we check every tile that leaves our shed, and we have seen every shortcut other suppliers take. This is the walkthrough we give customers who visit us in person, written down for the ones who cannot.

Walk into any slate factory worth its salt and you will see a row of tables. Not machines — tables. People sitting at them, one tile at a time, tapping, looking, flipping, sorting. That is the single biggest tell of a quality producer: the inspection line is staffed by humans, not replaced by a batch sample and a certificate.

We run that line in Jiujiang every working day. Customers visit, and the first thing we walk them through is not the quarry face or the CNC line — it is those tables. Because the question "is this slate any good?" is not answered by a brochure. It is answered by a repeatable process that catches the bad tiles before they go into a container.

Below is that walkthrough, broken into six stations. You can do the first three yourself with a sample in your hand. The last three require paperwork or a factory visit — but if a supplier cannot answer them, you have your answer about the slate.

Station 1 — Surface: What You See Before You Touch

Pick up a tile. Hold it flat, face up, at arm's length under good light. Tilt it slowly. What you are looking for is what every inspector in our shed looks for first: the surface tells you more than any datasheet.

Pass: a clean, natural cleft face with consistent colour across the tile. The cleavage surface — the flat plane the slate was split along — should be uniform, with a fine, slightly textured feel. Minor colour variation between tiles is normal and expected; slate is a natural stone, not a factory tile.

Fail: any of these on more than a handful of tiles in a sample of fifty:

• Cracks — hairline or open. Even a hairline crack will widen under freeze-thaw cycling. A cracked tile is a rejected tile, full stop.
• Veins or ribbon lines — bands of different mineral running across the face. These are zones of weakness; the slate will eventually split along them.
• Pyrite spots — small metallic inclusions that rust and bleed orange stains down the roof within a few seasons. Some Spanish and Chinese slates carry them; good producers sort them out.
• Surface scaling or delamination — the face is already flaking before it has been on a roof. That is a stone that will not last.

The tolerance we hold ourselves to: if one tile in fifty has a visible surface defect, the whole batch gets re-sorted. That is stricter than any standard requires — EN 12326 and ASTM C406 test performance properties, not cosmetic perfection — but a roof full of pyrite stains or hairline cracks is a roof that generates complaints, regardless of what the certificate says.

One thing that is not a defect: natural texture variation between the cleft face and a dressed edge. If you order smooth-dressed slate, the edges should be machine-trued. If you order natural cleft, the surface will have the organic roughness of hand-split stone. Both are correct — they are different slate types, not different quality levels.

Station 2 — Weight: The Cheapest Density Test You Can Do

Put two tiles of the same declared size on a scale. If one is significantly heavier than the other, one of two things is happening: the thickness is inconsistent, or the density is inconsistent. Both are problems.

Density matters because it is the umbrella property. A dense slate (2.7–2.9 g/cm³) absorbs less water, resists freeze-thaw better, and holds a higher flexural strength. A light, porous slate feels fine in your hand and fails on the roof. Weight is the proxy you can measure without a laboratory.

Here is a practical check with a single tile: a 500×250×5 mm slate from a quality quarry should weigh about 1.65–1.75 kg. If you are holding one that weighs 1.3 kg, it is either thinner than declared or the stone is unusually porous. Either way, ask why.

The numbers behind this: our Jiujiang slate tests at a density of 2.78 g/cm³ — comfortably in the middle of the good range, and consistent batch to batch because the quarry face is geologically uniform. You can read the full density and weight data on our slate weight guide and our published test results page.

Station 3 — The Ring Test: 19th-Century Tech That Still Works

This is the oldest quality check in the trade, and it is still the fastest. Hold a tile loosely by one corner — do not grip it tight, or you dampen the vibration — and tap the centre with a knuckle or a light tool. Listen.

A good slate rings. Clear, bright, sustained — like tapping a fine ceramic. The sound means the stone is dense, crack-free and homogeneous.

A bad slate thuds. Dull, short, dead. The sound means there is an internal flaw: a micro-crack, a vein, a soft zone. The tile looks fine on the surface, but the ring test catches what the eye cannot.

We do this on the inspection line for every tile. It takes two seconds per tile, and it catches the ones that pass visual inspection but would fail on the roof. Is it scientific? No. Does it work? Yes — well enough that the Welsh quarrymen relied on it for 200 years before any laboratory existed to validate it.

The modern version is the flexural strength test under EN 12326 or ASTM C406: a machine bends the tile until it breaks and measures the load. Our slate tests at a mean flexural strength that meets the S1 grade requirement. The ring test is the field version of that lab test — it will not give you a number, but it will sort the bad from the good before you spend money on a lab report.

Station 4 — Edges: Where CNC Makes the Difference

Flip the tile over. Look at the edges — all four. Run your finger along each one. What you are checking is the thickness tolerance: how close the actual tile is to its declared dimension, and how consistent that dimension is across the batch.

Why this matters on the roof: a roofer lays slate in courses, fixed to battens at a set gauge. If tiles vary in thickness by 2 mm, the courses will not lie flat. The roof will look wavy from the street, and the overlap (headlap) will be inconsistent — which means inconsistent weather resistance. A 0.5 mm tolerance installs cleanly; a 2 mm tolerance fights the roofer all day.

There are two ways to cut slate edges:

• Hand-split only. The splitter works to the cleavage, and the thickness follows the rock. Natural, but the tolerance is ±1.5–2 mm at best. Acceptable for random-width rustic roofs where variation is the aesthetic. Not acceptable for a uniform, architectural roof.
• CNC-dressed. After splitting, the tile goes through a CNC machine that trues all four edges to a programmed tolerance. Our CNC line holds ±0.5 mm on thickness and ±1 mm on length and width. That is the difference between a roof that lays flat and one that does not.

If you are buying uniform slate for a contemporary or heritage-match project, ask the supplier: "Are the edges CNC-dressed, and what is your thickness tolerance?" If the answer is "hand-split, about 1–2 mm either way," budget for a longer installation and a pickier roofer.

Station 5 — The Lab Report: Numbers That Do Not Lie

Stations 1 through 4 are things you can check with your hands and ears. Station 5 is the one that separates a supplier you can trust from one you cannot. Ask for the test report. Not a summary, not a claim, not a line that says "EN 12326 compliant." The actual report.

Two standards matter, depending on where the slate is going:

EN 12326 (European standard)

The report should give you at minimum these four numbers:

• Water absorption — the single most important figure. EN 12326 classifies slate by water absorption code: A1 (≤0.6%), A2 (0.6–10% with special testing) and A3 (above 10%, not suitable for roofing). A1 is the only grade you should accept for a roof expected to last decades. Our slate tests at 0.18% — well under the A1 threshold.
• Flexural strength — the breaking load. Expressed in MPa. The EN standard requires a minimum; S1-grade slate exceeds it comfortably. Low flexural strength means the tile cracks under foot traffic or hail.
• Freeze-thaw cycling — the test that simulates decades of winter. The tile is cycled between freezing and thawing hundreds of times. Our slate showed zero visible damage after the full cycle and maintained its flexural strength. That is the number that lets you sleep through a Minnesota January.
• Carbonate content — expressed as CaO percentage. Carbonates are the minerals that weather fastest under acid rain and UV. High carbonate (above 1–2%) means the slate surface will erode and the tile will thin over decades. Our CaO is 0.56% — low enough that surface weathering is negligible over a 100-year service life.

ASTM C406 (US standard)

ASTM C406 grades slate as S1, S2 or S3 based on water absorption, flexural strength and acid immersion testing. S1 is the highest grade and the only one we ship. S2 is acceptable for some moderate climates; S3 is not suitable for long-life roofing. The grade should be printed on the report, not just claimed in an email.

If you want the full breakdown of what each grade means, our slate grades guide walks through every code. Or you can read our own published test results — we put them online because a supplier who will not show you the report is a supplier with something to hide.

Red flag: if the supplier sends a report but it is undated, unsigned, from a quarry with a different name, or more than three years old, ask for a current one. Slate from different parts of the same quarry face can vary; a report from 2019 does not certify a batch produced in 2026.

Station 6 — Packing: What the Pallet Tells You About the Factory

The last station is not about the stone. It is about the culture. Walk to the packing area and look at how finished tiles are prepared for shipping. The packing tells you whether the factory cares about what happens after the container doors close.

Good packing:

• Tiles stacked face to face and back to back — the cleft faces are protected from rubbing against each other, which would dull the surface during 6,000 nautical miles of vibration.
• Foam or cardboard separators between layers — prevents edge chipping.
• Steel strapping (not plastic) on a wooden pallet — holds the load rigid under container movement.
• A breakage allowance packed separately — typically 2–3% extra tiles, clearly marked, so the installer has replacements on site without reordering.

Bad packing: loose-stacked tiles, no separators, plastic straps that snap in transit, no breakage buffer. If a factory cuts corners on packing, it cuts corners on inspection — the mindset is the same.

We pack every pallet as if the customer is going to open it in front of us — because eventually, they do. The packing is the last quality check, and it is also the first thing the customer sees when the container arrives. A broken, dusty, chipped pallet sends a message that no test report can undo.

The Scorecard: Putting All Six Together

One station in isolation does not tell you much. A supplier with a great lab report but a sloppy inspection line will ship defects. A supplier with beautiful surfaces but no test report is selling you aesthetics, not performance. The point of the six stations is that they reinforce each other:

Station	Checks the risk of...	Who can do it
1. Surface	Visible defects, pyrite, cracks	Anyone with a sample
2. Weight	Low density, inconsistent thickness	Anyone with a scale
3. Ring test	Internal flaws, weak zones	Anyone with two hands
4. Edges	Poor tolerance, installation problems	Anyone with a calliper
5. Lab report	Absorption, strength, freeze-thaw, carbonate	Supplier must provide
6. Packing	Transit damage, quality culture	Factory visit or photos

Here is the honest summary: if a supplier passes all six, the slate will perform. If they pass the first four but not the fifth, the slate might look good and still fail in ten years. If they pass the fifth but not the first four, the certificate is real but the batch control is not — and you will get good stone mixed with bad.

The reason we run all six, every day, on every tile, is that the shortcut is always available and always expensive. Skip the ring test and a cracked tile ships. Skip the lab report and a whole container goes to a climate it cannot survive. Skip the packing and the installer opens a pallet of chips. Each shortcut saves a few minutes and costs a customer.

If you are evaluating suppliers and want a benchmark to compare against, request a sample from us — the form is here — and run the six stations yourself. Then ask the same questions of every other supplier on your list. The slate that passes all six is the slate that belongs on your roof.

Frequently Asked Questions

Related Reading

• Jiujiang Slate Test Results — the actual EN 12326 and ASTM numbers behind every claim in this article, published and current.
• Roofing Slate Grades: S1, S2 & T1–T3 Explained — how to read the grade codes on a lab report and what each tier actually means on the roof.
• 8 Properties of Natural Roofing Slate — the science behind the six checks: density, absorption, flexural strength and more, with the numbers.
• Roofing Slate Buying Guide & Checklist — the full procurement checklist, from samples to container loading, for first-time importers.
• Is Natural Slate Suitable for Every Climate? — how to match the quality checks in this article to the climate where the slate will live.