Pebbles, Decorative Glass & Volcanic Rock: Order by Volume, Not Weight

The three numbers that decide whether you got a good deal on decorative stone

If you're buying pebbles, decorative glass, or volcanic rock by the tonne, pallet, or bulk bag, stop comparing suppliers by price per tonne and start comparing them by price per square meter of installed coverage. Those are two different numbers, and the gap between them is where B2B buyers quietly lose money.

Three data points do most of the work: bulk density, fines content, and the actual sieve range. Get all three in writing before you commit to an order. Not a sample bag from the rep's trunk — the numbers from the run that's actually going on the truck.

Here's the figure that surprises people: one tonne of scoria fills roughly twice the volume of one tonne of marble chips. Carry a coverage number from one product over to another and you'll be off by 50% or more. You won't find out until a crew is standing in a half-finished bed waiting on a second delivery.

Why you should trust someone who got this wrong repeatedly

I've been handling decorative aggregate and hardscape material orders for nine years — first as a buyer for a landscape contractor, now in procurement for a garden supply distributor. I've made and documented 23 significant ordering mistakes in that time. Roughly $14,000 of them came out of our margin in reorders, extra freight, and crew hours we couldn't bill.

Three of those mistakes taught me almost everything I use today.

March 2021. I ordered 14 tonnes of 20–40 mm white marble chips for a courtyard, using a coverage number I'd carried over from a lava rock job the year before. Marble chips are dense. Lava rock isn't. We ran about 40% short. The reorder came from a different production batch and didn't match. Total damage: $1,850 in extra freight plus six days of a crew working around a hole in the plan.

September 2022. We shipped 'washed' 10–20 mm pebbles straight into a water feature without rinsing them. The supplier's definition of washed and mine weren't the same thing. Three weeks of cloudy water, an unhappy client, and $1,200 in labor to drain, rinse, and refill — for something an afternoon of prep would have prevented.

Q1 2024. Dyed red rock for a south-facing display garden. It looked great for eight months. Then it didn't. It faded to a washed-out pink, and the top-up batch we ordered was noticeably darker because the dye lot had changed. We replaced 60 bags at our own cost. That's the mistake that made me start keeping reference bags on a labeled rack by the loading dock.

It took me four years and somewhere around 60 bulk orders to internalize this: the spec sheet tells you what a supplier intends to ship; the as-delivered condition tells you what you actually bought. Those two things are usually close. 'Usually' is where the money goes.

Where the money actually leaks

Sizing: 20–40 mm is a range, not a promise

Nominal size ranges are real, and some percentage of undersize and oversize material is normal — that's how screening works. What isn't normal is when the undersize fraction climbs past roughly 10–15%, because fines are the part that clouds water, stains paving, and fills the voids that make a drainage layer function.

The standard test behind those numbers is ASTM C136, sieve analysis of fine and coarse aggregates. Most decorative suppliers won't have a gradation report — they aren't selling to concrete spec — but anyone selling into commercial hardscape should be able to tell you which screens they run product on. If they can't answer that, that's your answer.

Two more sizing traps worth knowing. First, flat and elongated particles: a stone that's 30 mm across and 9 mm thick is technically in spec and useless for a walking surface, because it packs into a mosaic of edges instead of a stable layer. Second, blends. Mixed-size product compacts better than a single size in some applications, but blends are also where odd lots get sold as 'natural mix.' Ask for the ratio.

Density: coverage figures don't transfer between products

Bulk density decides how far a tonne or a bulk bag actually goes. The spread across decorative materials is wider than most buyers assume:

• Marble and granite chips: roughly 1,400–1,700 kg/m³, sold almost always by weight
• River rock and oval pebbles: roughly 1,400–1,600 kg/m³
• Scoria and lava rock: roughly 600–900 kg/m³, often sold by volume instead, because bagging it by weight is impractical
• Decorative glass: roughly 1,300–1,600 kg/m³, depending on crushed vs. tumbled

So when one supplier quotes lava rock per tonne and another quotes it per cubic meter, you're looking at two numbers you can't compare without the density. I've watched two buyers argue about this for an hour. Only one of them was right, and it wasn't the one holding the smaller number.

Sanity-check method: a tonne of dense 20–40 mm stone at 50 mm depth covers roughly 9–13 m². The same tonne of scoria at the same depth goes closer to 20–27 m². If a quoted coverage figure lands well outside that, the supplier is either assuming a different depth than you are, or guessing.

One more weight-related point: moisture. Washed product arrives wet, and wet weight is weight you paid for. On a single bag it's trivial. On a full load it isn't. Nobody's cheating you — that's just how bulk material works — but it belongs in your per-square-meter math.

Washed vs. unwashed: the least standardized word in the category

Washed has no shared definition for decorative product the way it does for concrete aggregate. For one supplier it means run through a spray bar. For another it means it rained before bagging. Both will print it on the invoice.

Where it bites: water features, indoor gravel, and anything light-colored. Unwashed stone carries dust that clouds water for days, stains pale grout, and settles into every corner of a planter. You can rinse on site — mesh screen, hose, patience — but it's slow and genuinely miserable in cold weather. I'd rather pay a premium for product that's actually been washed than put two people on a rinse detail for a day.

Color: dye lots, fading, and the word permanent

Colored rocks for landscaping fall into three buckets, and each one fails differently.

Dyed. Color is soaked or sprayed into porous stone. Vivid out of the bag, and it fades. How fast depends on UV exposure, and most dyed landscaping rock carries no UV-stability claim at all. When a supplier says the color is permanent, ask for the test method. There isn't a widely used standard for this product category, which is exactly why the claim gets made so casually.

Coated. A resin or paint layer sits on the surface. Holds color better, chips more easily — and a chipped coated pebble against a light surround is very visible. Fine for beds nothing touches. Poor for paths.

Natural. The stone's own mineralogy. No fading, no claims needed, and a narrower palette: grays, tans, browns, blacks, off-whites. If the design calls for cobalt blue, you're buying dyed material, and you should plan for replacement in a few years rather than pretend otherwise.

Then there's dye lots. Colored product varies between production runs, sometimes enough to see across a walkway. If a project needs more than one pallet, order it all at once and keep one unopened bag as a physical reference. Our reference rack has prevented two customer escalations this year alone.

Decorative glass: the difference is edge finish

Crushed glass has edges. Tumbled glass doesn't. That's the whole distinction, and it matters more than the color range.

Crushed glass is sharp enough that I won't specify it anywhere people walk barefoot, and I'd keep it away from pool decks and play areas entirely. Tumbled glass rounds the edges mechanically, costs more, and is what you want anywhere someone might sit down or walk in sandals.

If glass is going into or near a fire feature, it needs to be tempered glass made for that purpose. Ordinary decorative glass — including a lot of the recycled material sold for landscaping — can shatter under thermal shock. Ask specifically for tempered fire glass, and don't accept 'it's glass, it'll be fine.'

Volcanic rock: three materials, one trade name

Lava rock gets used as a catch-all for products that behave very differently from each other.

Scoria is the porous, lightweight material that floats for a second or two before it sinks. Pumice is lighter and softer, and it breaks down under foot traffic — not what you want on a path. What's sold as lava rock is usually scoria or a scoria mix, and it earns its place as a drainage layer, a mulch alternative in xeriscape beds, and a filler where you need volume without much weight.

Two things nobody mentions until you've handled a pallet. Scoria is dusty — it arrives coated in fine material that gets on hands, gloves, and anything pale nearby (note to self: unload it upwind of the white marble). And some of it is genuinely sharp. Not glass-sharp, but you wouldn't want to crawl across a bed of it. Great for beds nobody walks on. Think twice for indoor gravel trays and paths.

Indoor gravel: dust first, chemistry second

Indoor gravel usually runs 3–8 mm, and the problems are different from outdoor work: dust on hard floors, dust in the air for the first week, and the question of whether the stone is chemically inert.

For anything near a fish tank or a planted water feature, test before you buy in quantity. A few drops of vinegar on a sample will fizz if the stone is carbonate-based — limestone, marble, shell — which means it'll push pH up in water. Marble chips are exactly the material that looks best in that role and behaves worst. If you don't know what you've got, ask for the stone type in writing, not a trade name. That's a case where I genuinely wouldn't guess.

For dry indoor use — planters, trays, decorative beds — the rules are simpler: rinse it, let it dry, and use a liner so you aren't pouring water through dust into whatever sits underneath.

When this advice doesn't apply

If you're buying six bags for a balcony, almost none of this matters. Volumes are tiny, freight is trivial, and rinsing a bag of pebbles in a bucket takes ten minutes. The math above is for orders where reorders, freight, and crew time cost more than the material itself.

If you're buying retail rather than wholesale, you often can't get density figures or screen sizes, because the retailer doesn't have them either. That's not a red flag — that's how packaged retail works. In that case the checklist shrinks to one step: buy a single bag, open it, look at it, then buy the rest. That's genuinely enough.

And I can only speak to our context — a mid-size distributor buying pallets and part-loads for regional projects, mostly from repeat suppliers. If you're importing full containers, there are pre-shipment inspection and moisture-certification issues I've only brushed against, and I wouldn't pretend to know them well.

One thing I still don't understand: why two suppliers running nominally identical screens produce visibly different product. My best guess is feed rate and screen wear, but I've never confirmed it. If you've figured that one out, I'd like to hear how.