We Saved $184K on FCC Catalyst. Then the MAT Results Came Back.

January 2023: A Renewal Window I Thought I Had Under Control

January 2023. Our FCC unit was about nine months out from a scheduled turnaround, and the equilibrium catalyst contract was up for renewal. I handle catalyst and chemical spend for a 68,000 bpd refinery — roughly $2.4 million annually on FCC catalyst alone, plus another $1.8 million spread across hydrotreating, reforming, and water treatment chemicals. Six years in the seat. Forty-some vendors negotiated. Every invoice logged in our cost tracking system, going back to my first week on the job.

So when I say I thought the renewal was routine, understand that I wasn't being naive. I had a spreadsheet with three years of price history, a TCO model I'd rebuilt twice, and a policy requiring minimum three competitive quotes for anything over $50,000. I did the work.

What I didn't do was read the specifications closely enough. That's the whole story, and it's worth telling because the mistake cost us — conservatively — $190,000 in a single contract cycle.

Three Quotes, One Obvious Winner

We run about 350 tons of fresh catalyst per year. The incumbent supplier quoted $6,400 per ton, delivered, with a June-to-June pricing window. Second bidder came in at $5,880 — an 8% cut, roughly $182,000 in annual savings. Third bidder was $5,640 but needed six weeks of lead time and had no rail siding access, which meant we'd be eating truck freight on every delivery.

On paper, this wasn't close. The second bidder had the price, the delivery window, and a decent service record on our hydrotreater catalyst account the previous year. Our procurement policy requires three quotes; I had three. The TCO model said yes. My manager asked one question — "are the specs identical?" — and I said yes, because the data sheets looked identical in the ways I knew to check.

That's the part I keep coming back to. They looked identical. I was checking the wrong lines.

Week Three: The MAT Numbers

Here's the thing about FCC catalyst that took me too long to internalize: the price per ton is about the least useful number in the whole transaction. What actually drives your economics is activity, attrition resistance, and unit cell stability — and those live in test results, not on a quote sheet.

We pulled a composite sample three weeks into the first delivery and sent it to an independent lab running ASTM D3907 (microactivity test, MAT) and ASTM D5757 (air-jet attrition). The contract spec sheet said MAT ≥ 70. The lab came back at 66.4 on the first composite, 67.1 on the second.

Below spec. Not catastrophically — but enough that our unit's conversion dropped more than I'd budgeted for, and enough that fresh catalyst addition rate had to go up to hold activity in the circulating inventory.

Then the attrition data came back worse than the activity data. Fines generation up, cyclone loading up, and our stack opacity readings started nudging the permit limits on high-throughput days. That's a red flag you don't ignore, because the fix for opacity excursions isn't a procurement fix. It's an environmental compliance conversation.

The Specification Sheet Said "Standard Grade"

Here's where it got embarrassing. I pulled both data sheets side by side and started reading the fine print on rare earth content. Our incumbent's product ran 2.4–2.6% RE₂O₃. The new supplier's product ran 1.8%.

Both documents used the phrase "standard grade." Both were, technically, telling the truth — there's no industry-wide definition of "standard" for equilibrium catalyst rare earth loading. I said standard grade. They heard standard grade. We meant different numbers, and neither of us caught it until the lab did.

Looking back, I should have written the spec as a numeric range, not a word. "RE₂O₃: 2.3–2.6 wt%" is a spec. "Standard grade" is a conversation with a lawyer waiting to happen.

The Rescue, and What It Cost

We had nine weeks to the turnaround window. Going back to the incumbent meant an expedited order at a 6% premium plus air freight on the first two rail cars, because their June production slot had already been reallocated when we walked. That premium ran $108,000.

Independent lab testing on both incoming and in-unit samples for the rest of the cycle: $14,000.

Increased fresh catalyst addition to hold activity through Q3: $47,000.

And the piece that still makes me wince — the internal yield-loss estimate from running below-spec activity for roughly eleven weeks. Our process engineering group put the light-cycle oil and gasoline yield impact north of $200,000. I've seen the model. I believe the number.

Net it out against the $182,000 in unit-price savings and we lost somewhere around $190,000 on the deal. The "cheap" bid was the most expensive option on the table, and it wasn't close.

Bottom line: the quote sheet tells you what a supplier wants to charge. The spec sheet tells you what you're actually buying. Only one of those numbers shows up in your budget variance report, and it's the one nobody put on the quote.

What Changed After That

Three things, and I've held to all of them for the last two contract cycles.

First, no quote without test data. Any FCC or refinery catalyst bid over $75,000 now has to arrive with ASTM D3907 and D5757 results from an independent lab, dated within the previous six months, on the specific production lot being offered. Not a generic product data sheet. Not a certificate of analysis from the supplier's own QC lab. Third-party, lot-specific, dated. Two suppliers have walked away from that requirement. That tells me something useful about both of them.

Second, numeric specs, always. No "standard," no "premium," no "high-activity" in any document I sign. If a parameter matters to the unit, it gets a number and a tolerance. It costs me maybe two extra hours per contract to write it that way. It cost me $190,000 not to.

Third, and this one took longer to internalize: I stopped treating catalyst sourcing as a price comparison exercise. The product quality feeds directly into what our refinery can put out the door, and what our customers see when they open a tank car. Batch-to-batch variability doesn't stay in the catalyst inventory. It shows up in the product slate, and downstream of that, in whether our offtake partners think we're a reliable supplier or a headache. The money I "saved" on the unit price got paid back by the production team in yield they couldn't recover and in conversations I'd rather not have again.

Here's what I tell new buyers on my team now: your cheapest line item is the one where the supplier can prove, with dated third-party data, that the product does what the spec says. Everything else is a price. A price is not a cost.

I still keep the original quote comparison from January 2023 in a folder on my desktop. It's the most useful procurement document I own, and it's useful precisely because every number on it was wrong in a way I couldn't see at the time.