6-Step Lab Equipment Approval Checklist: What a Quality Inspector Checks Before Signing Off

Over the last seven years—six, if you count my first informal year—I've been responsible for accepting and rejecting analytical equipment at a mid-sized materials testing lab. I review every instrument before it goes into production: roughly eight to ten major systems a year, plus a stack of smaller tools. I've reviewed maybe 40 instrument deliveries in total. Perhaps 37; I'd have to check the log. In 2024, we rejected roughly 12% of first deliveries for documentation and spec mismatches. Not because the vendors were dishonest. Because the buying team skipped the checklist.

If you're specifying any lab instrument, use this before you issue a PO. It works for a $2,000 digital micrometer and a $150,000 Thermo Fisher Raman microscope rental.

Step 1: Force the vendor to prove the application match

The biggest mistake is comparing spec sheets.

If you're buying a Thermo Fisher fluorescence microscope, don't just compare magnification and camera pixels. Ask for images of your specific fluorophores, your sample thickness, your signal-to-noise conditions. A fluorescence microscope that excels at fixed cells can be a poor fit for live-cell work or tissue sections. In 2023, a vendor quoted us a microscope with impressive specifications. The first test slide came back with so much autofluorescence we thought the filter cube was wrong. It wasn't. The system was fine for brighter signals—just not for ours.

Write into the PO that acceptance requires a demonstration using your sample type. If the vendor refuses, walk away.

Step 2: Run the rental math before you buy

Here's something vendors won't tell you: rental isn't just for short-term needs. It's a due diligence tool.

For high-cost systems like a Thermo Fisher Raman microscope rental, the rental fee can be treated as the price of information. We rented one for 90 days before committing to a purchase. During that window, our operator discovered that the 532 nm laser was ideal for the samples we thought we'd need—but the 785 nm configuration mattered more for the next project. Had we bought the first configuration, we'd probably have paid for a wavelength change later.

Ask whether rental payments can be credited toward a future purchase. Some programs allow it. Even if not, the cost of avoiding a wrong $100k decision is worth it. Price matters, but the worst price is the one you pay twice.

Step 3: Verify traceability—a digital micrometer is still a micrometer

Let's talk about the small tools, because they cause the most arguments.

A digital micrometer is not automatically accurate just because it has a digital readout. In plain terms, the 'micrometer digital readout' only tells you what the sensor measures; it does not prove the measurement is right. I have rejected micrometers because the calibration certificate had expired. Seems minor. In a QA lab, minor is how measurement error gets into the record.

When a digital micrometer arrives, check three things:

• Resolution and accuracy per the manufacturer's spec sheet—and make sure the accuracy claim includes repeatability.
• Calibration certificate: NIST-traceable or ISO 17025-accredited, with the uncertainty budget. Don't accept 'calibrated' without the document.
• Data output compatibility. A digital micrometer with USB or Bluetooth is only useful if the software can export into your LIMS format.

Also check the IP rating if it's going onto a factory floor. A digital micrometer with IP54 can handle dust and splashes; IP40 won't. The catalog page tells you this in fine print. Read it.

Step 4: Run liquid chromatography qualification with your own standards

When a liquid chromatography system arrives, the vendor will run a test and show you a beautiful chromatogram. That's good. It's also worthless unless those peaks are your compounds.

For an HPLC or UHPLC system, run the same acceptance protocol every time:

1. Injection precision: six replicate injections, calculate %RSD. For most QC work, you want %RSD under 0.5% for retention time and under 1% for area.
2. Detector linearity across the concentration range you actually use.
3. Carryover: run a blank after your highest standard and look for a ghost peak.
4. Baseline noise and drift under your worst-case gradient.

If your deliverable acceptance criteria aren't in the PO, the vendor is not obligated to meet them. Write them into the contract. The two hours this takes on the front end saves the two-week dispute later. I've lived that dispute. It's not fun.

Step 5: Bench-test durability before standardizing on a sensor

This one applies more to the equipment group than the analytical lab, but it's the same decision discipline.

When we needed to standardize photoelectric sensors for a packaging line, the durability of ifm photoelectric sensors vs others became the deciding factor. On paper, two competing options looked similar. The difference was failure rate over time.

We ran a side-by-side trial: 20 sensors from each supplier on the same vibrating frame, 400,000 cycles, with dust and temperature swings. The ifm sensors finished with one failure. One of the other brands had four. The third had six.

The ifm sensors cost more upfront. But the cheap ones caused line stoppages that cost more in lost output than the entire sensor purchase. Now my rule is: if a component can stop your line, don't choose it on price alone. Test it first.

I'm not saying every cheaper sensor is bad. I'm saying you don't know until you run it in your conditions. That data is worth more than the vendor's marketing number.

Step 6: Check service and support before you need it

A service plan might look like an unnecessary line item. It isn't.

Before signing, ask these questions:

• What is the response time for a critical instrument down? 24 hours? 72 hours? Put it in writing.
• Are spare parts stocked locally or does every repair wait on international shipment?
• Who provides the training? Is it included or billed per day?

We bought an instrument once with the lowest quote and the most optimistic delivery timeline. The vendor's after-sales support was unresponsive. What we saved on purchase price, we lost in instrument downtime and rushed freight. The math didn't work. Total cost, not invoice price, is the only honest comparison.

Common Mistakes That Still Happen

Here are the errors I see even from experienced buyers:

• Forgetting installation requirements. A Raman microscope needs chiller clearance; a liquid chromatography system needs proper exhaust and solvent storage. Discover this after delivery and you're paying for a second installation visit.
• Not budgeting consumables. Filters, columns, calibration standards, sensor replacement parts. The instrument might be capital spend, but the consumables are operating expense and they're predictable. Budget them.
• Accepting paperwork without checking lot numbers. In Q1 2024, we received a shipment of standards where the certificate of analysis listed a different lot number from the bottle label. It was an honest labeling error, but we rejected it anyway because we couldn't verify the value. Documentation is part of the deliverable.

The numbers will tell you a lot. Your gut, though, is also data. Listen when a sales engineer can't answer a basic question, when the quote has too many conditional notes, when the rental terms feel vague. Every expensive mistake I've made had a quiet warning sign. The checklist helps you notice it before you sign.