Written and maintained by CASRAI Editorial Board
Last updated
A radiation program that treats “the survey meter is calibrated” and “the survey meter passed today’s response check” as the same fact will fail an inspection on the gap between them. They are two different obligations, done by two different people, on two different schedules, against two different standards — and only one of them can actually be satisfied by sending the instrument out once a year.
Two obligations, not one
The confusion is understandable because both procedures point a check source at the same detector. What they establish is different:
- Calibration is an external, NIST-traceable determination of the instrument’s actual response across its full measurement range, performed by a calibration laboratory (in-house or vendor) against a known, characterized source geometry. It produces a calibration certificate and, typically, a correction factor.
- Response check (also called a source check or operational check) is an in-house, pass/fail confirmation that the instrument still responds the way it did at its last calibration — not a re-measurement of its accuracy. It’s performed by the user, not a calibration lab, and takes under a minute.
A response check that passes does not mean the instrument is in calibration; it means nothing has obviously broken since the last time it was. Only the calibration itself establishes accuracy.
What 10 CFR 20.1501 actually requires
10 CFR 20.1501 is the general NRC survey requirement: licensees must make surveys reasonably necessary to comply with Part 20 and to evaluate radiation levels, and the instruments used for those surveys must be calibrated periodically. Read carefully, 20.1501 does not itself specify a calibration interval — that number comes from elsewhere, which is the single most common source of confusion in this area.
Where the “annual” interval most labs work to actually comes from: 10 CFR 35.61, which applies specifically to medical-use licensees, requires survey instruments used to demonstrate compliance be calibrated before first use, at intervals not to exceed one year, and after any repair that could affect calibration — calibrating at two separated points (or one point per decade) on each scale up to 10 mSv/hr (1,000 mrem/hr) using a radiation source. A broad-scope academic or research license is not bound by Part 35 unless it also holds medical-use authorization, but most institutional radiation safety programs adopt the same annual interval as a license condition or internal policy anyway, because it is the interval NRC inspectors are most familiar with and the one most calibration vendors default to. If your institution’s specific license or radiation safety manual states a different interval, that document controls — not this general convention.
The calibration standard: ANSI/IEEE N323AB-2013
ANSI/IEEE N323AB-2013, American National Standard for Radiation Protection Instrumentation Test and Calibration, Portable Survey Instruments, is the current consensus standard NRC-licensed calibration facilities test against. Two things about it matter for a program deciding how to structure its own calibration and check program:
- It allows control-chart-based interval extension. Instead of a flat annual interval for every instrument regardless of behavior, N323AB permits a facility to track an individual instrument’s response-check history on a control chart over time and extend its recalibration interval if that history shows stable performance — a data-driven alternative to the fixed calendar interval, provided the underlying license condition doesn’t already fix the interval by regulation (as 35.61 does for medical-use licensees).
- It reclassifies the source response check as an “environmental check,” not a calibration step — formally separating it from the calibration procedure itself, which is exactly the operational distinction this page opens with. A check source, per N323AB’s own definition, is “a radioactive source, not necessarily calibrated, which is used to confirm the continuing functionality of an instrument,” used in the same source-to-detector geometry as the original calibration.
For an ISO/IEC 17025-accredited calibration function specifically, the calibration certificate itself needs to show NIST traceability, the as-found and as-left response at each tested point, the correction factor applied (if any), the source used and its certified activity/geometry, and the technician and date — the elements an accreditation auditor checks for, not just a pass/fail stamp.
The daily response check
Separately from annual calibration, most institutional radiation safety programs require every portable survey meter to pass a response check the first time it is used each day, using the same check source and geometry every time. This is standard, published practice at university radiation safety programs (Princeton and UC Berkeley EH&S guidance both state it explicitly) rather than a universal regulatory mandate under 20.1501 itself — check your own institution’s radiation safety manual for the exact frequency it imposes, since some programs require it before each day’s use and others before each discrete survey.
Because N323AB-2013 dropped a specified numeric acceptance criterion for the response check, most programs still use the threshold from the older ANSI N323-1978 standard as their working pass/fail line: if the check-source response differs from the value recorded at the instrument’s original calibration by more than ±20%, the instrument is pulled from service and sent for recalibration or repair — it is not used for surveys or compliance measurements again until it passes. Some programs set a tighter internal action limit (commonly ±10–15%) as an early-warning threshold before the hard ±20% failure line, logging the trend rather than waiting for an outright fail.
A response check that fails does not necessarily mean the instrument is inaccurate across its whole range — it means something changed (battery, detector, source decay since the last check, physical damage) that has to be resolved by recalibration before the instrument can be trusted again. Treating a failed response check as ‘close enough, recalibrate at the next annual cycle’ is the actual compliance gap this two-tier system exists to catch.
Calibration vs. response check, side by side
| Calibration | Response (source) check | |
|---|---|---|
| What it establishes | Instrument accuracy across its measurement range | Nothing has changed since the last calibration |
| Who performs it | Calibration laboratory (in-house or vendor), NIST-traceable source | The instrument’s user |
| Frequency | Typically annual (35.61 for medical-use licensees; program policy otherwise), or control-chart-extended per N323AB | Before first use each day (typical program policy) |
| Governing reference | 10 CFR 35.61 / license condition; ANSI/IEEE N323AB-2013 test procedure | ANSI/IEEE N323AB-2013 (‘environmental check’); ±20% threshold from ANSI N323-1978, still widely used |
| Failure outcome | Instrument out of service until recalibrated | Instrument out of service until recalibrated (a failed check is not itself a repair) |
| Documented as | Calibration certificate on file | Daily response-check log entry (date, source ID, reading, pass/fail) |
Documentation an inspector or accreditation auditor will ask for
Keep the two records separate, since they answer different questions:
- The calibration certificate for every instrument currently in service — showing the calibration date, next-due date, the points tested, as-found/as-left readings, and traceability to a NIST-traceable source.
- The daily response-check log — a running record (paper or electronic) of every check performed: date, instrument ID, check source used, reading obtained, and pass/fail against the program’s acceptance criterion. A missing or gapped log is a common inspection finding even when every individual instrument’s calibration is current, because it’s the only objective evidence that the daily-use requirement was actually followed rather than just written into the safety manual.
Both records tie into the same underlying obligation as occupational radiation dose limits under 10 CFR 20: a survey meter is only useful evidence of dose compliance if its readings are trustworthy, and trustworthiness is exactly what the calibration-plus-response-check pair is built to demonstrate.
Frequently asked questions
Can a passed daily response check substitute for annual calibration?
No. A response check only confirms the instrument’s response hasn’t drifted since its last calibration; it does not re-establish accuracy across the instrument’s range. Only a full calibration against a NIST-traceable source does that.
What happens if a survey meter fails its daily response check?
It is taken out of service immediately and not used for surveys or compliance measurements until it has been recalibrated or repaired and passes a fresh response check. Continuing to use it ‘until the next scheduled calibration’ is the gap this two-tier system exists to prevent.
Does 10 CFR 20.1501 specify a one-year calibration interval?
Not directly. 20.1501 requires periodic calibration without naming an interval; the widely-used one-year figure comes from 10 CFR 35.61 (which applies specifically to medical-use licensees) and has been adopted as common practice, license condition, or institutional policy well beyond that specific population.
What acceptance criterion should a response check use?
ANSI/IEEE N323AB-2013, the current calibration standard, no longer specifies one. Most programs still apply the ±20% threshold from the older ANSI N323-1978 standard as their working pass/fail line, since no current standard has superseded it with a different number.
For the broader regulatory structure this sits inside, see what a Radiation Safety Officer does, the Radiation Safety Committee (RSC), ALARA time, distance, and shielding practices, how dosimetry badges work, and radioactive waste disposal in the lab, or return to the lab compliance hub for the rest of this cluster’s coverage.
