How to choose a radiation calibration vendor

Calibration vendor selection is one of the most consequential procurement decisions an RSO makes — and one of the most underexplained in the literature. Send your survey meters to the wrong lab and you can end up with calibration certificates that don't actually cover your use case, instruments that come back out of spec, or worse: a regulator citation during your next inspection because your traceability chain has a broken link. This guide walks through what actually matters.

Accreditation is the floor, not a differentiator. ISO/IEC 17025 accreditation should be considered a minimum baseline for any calibration lab you'd send a regulated instrument to. But accreditation alone isn't enough — every accreditation has a *scope*, and that scope determines what the lab can and can't certify. Many labs are accredited for exposure rate measurements but not for tissue-equivalent dose. Others are accredited for gamma but not for beta. Some are accredited for low-energy photon work but not high-energy. Ask for the lab's Scope of Accreditation document and read it. If your survey meter use case isn't explicitly covered, the certificate they hand back doesn't legally certify what you need it to certify.

Traceability is the substance behind the certificate. NIST traceability — the audit trail from your instrument back to a national standard — should be documented through an unbroken chain of comparisons. Reputable labs participate in NIST round-robin proficiency tests and NVLAP measurement assurance programs. Ask for the lab's most recent proficiency test results. Labs with nothing to hide will share them; labs that hedge are signaling a problem. A 5 percent expanded uncertainty (k equals 2) on exposure rate is reasonable for most field instruments; uncertainties significantly higher than that should make you ask why.

Turnaround is a capacity signal. Five to ten business days is the industry standard for routine calibration. If a vendor quotes three-week turnaround as their normal lead time, that's a capacity problem, not a feature. Worse, it usually means your instruments will sit on a shelf for two of those three weeks waiting for someone to touch them — which raises questions about how they're stored, how they're handled, and whether your calibration is happening on the dates printed on the certificate or actually three weeks earlier. Ask for the typical time-in-lab vs total-turnaround breakdown.

Source variety determines what they can test. A lab with only Cs-137 and Co-60 sources can calibrate at a limited set of energies. If you use NORM survey meters, low-energy diagnostic dosimeters, or specialty instruments like neutron monitors, ask explicitly which sources they have and at what activities. Calibration at the wrong energy produces a certificate that overstates the instrument's accuracy for your actual use case.

Repair capability matters more than vendors admit. Roughly 15 to 20 percent of instruments sent for routine calibration fail to meet spec and need repair. Some labs do repairs in-house; others ship them out to the manufacturer or third-party repair shops. In-house repair compresses your total downtime by 1 to 3 weeks. Ask what their failure rate looks like across the instruments they handle, and what their repair-and-recal cycle looks like for the brands you use.

Calibration scope creep is real. Some labs will calibrate the energies and ranges you actually need. Others run their standard battery — which may include points outside your use case (wasting your money) or skip points within it (leaving you with a partial certificate). Tell them exactly what energies and ranges you need certified, and have them quote against that scope rather than their default test sheet.

Document control habits separate professionals from amateurs. Reputable labs send the certificate as a controlled document with a serial number, lab ID, technician initials, calibration date, due date, environmental conditions, and pre/post-cal readings for each test point. If you get a one-page certificate with no pre/post numbers and no environmental conditions, ask why — and seriously consider going elsewhere.

Geography matters, but not how you'd think. Local labs are tempting for the obvious shipping logistics, but the calibration ecosystem is national. A lab three time zones away that's truly excellent will outperform a local lab that's adequate. The shipping cost differential is rarely more than 200 dollars per instrument round-trip, which is noise relative to the cost of bad calibration. Pick the right lab first; optimize shipping second.

Pricing tells you something, but not everything. A typical 0 to 10 R/hr survey meter calibration runs 150 to 300 dollars in 2025-2026, depending on energy points and accreditation scope. Quotes significantly below that are either running a stripped-down protocol or loss-leading to build market share. Quotes significantly above are either offering specialty work or overcharging. Get three quotes for the same explicit scope and compare.

The contract terms you'll wish you'd negotiated. Annual contracts with multi-instrument volume discounts are standard. Ask for: same-day initial inspection reports (so failed instruments get triaged immediately rather than waiting in line), expedited turnaround for emergency cases, loaner instruments for critical applications, and explicit pickup/return shipping responsibility. Most labs will accommodate these; many won't volunteer them.

The questions to ask before signing. Beyond pricing and turnaround: Can I see your scope of accreditation? What's your typical pre/post-cal failure rate? Do you do repairs in-house or ship out? What energies do you offer and at what activity levels? Can you handle (specific brand/model) instruments? Do you provide loaners? What's your latest NIST round-robin result? What's your data retention policy if I need historical certificates 5 years from now? A vendor who answers all of these without hesitation is one you can rely on.

Red flags that should end the conversation. Reluctance to share scope of accreditation. Vague answers about proficiency testing. Standard turnaround quoted at 3+ weeks. Pricing more than 30 percent below the market median. No clear contact for technical questions vs sales questions. A pattern of certificates with missing fields. Any of these should make you keep looking.

For most facilities, the right answer is one annual contract with a single reputable national lab that covers all your routine work, plus one specialty lab on retainer for the unusual instruments that the primary vendor can't handle. That structure gives you the consistency that makes audits painless and the specialty access that keeps your unusual instruments functional.