Survey-Meter Dead Time and Count-Rate Losses

Every radiation counter is briefly blind after each event it records, and at high count rates those lost counts add up to a reading that understates the true field — sometimes dangerously. For one class of instrument the displayed rate not only lags the truth but actually falls as the real field rises, so an intensely contaminated surface or a hot therapy source can read deceptively low. Recognizing dead time, correcting for it, and knowing when an instrument has saturated are fundamental radiation-safety skills.

Introduction

Dead time (also called resolving time) is the minimum interval that must separate two events for the instrument to register both. It arises from the detector physics — the time an ionized gas takes to recover, or a scintillator and its electronics take to process a pulse — and from the counting electronics that follow. Because radioactive decay is random, some events always arrive within that interval and are lost. At low rates the loss is trivial; near saturation it dominates the measurement. ^{1, 2}

This article explains the two standard dead-time models, derives the count-loss corrections, shows how to measure an instrument's dead time, and connects the physics to real radiation-protection decisions. DRPS supports facilities with this analysis through its radiation safety officer and medical physics consulting services across Florida, Maryland, Virginia, Washington DC, California, Nevada, Pennsylvania, New York, New Jersey, and Delaware.

Topic Explanation

Why counters miss events

A gas-filled or scintillation detector does not respond instantaneously. After an ionizing event triggers a pulse, the detector and its electronics need time to recover before they can cleanly register the next one. During that recovery, a second event may be lost entirely, merged into the first (pulse pile-up), or only partly registered. The practical upshot is that the observed count rate $m$ is always less than or equal to the true event rate $n$ , and the difference grows with rate. ^{1, 2}

Two idealized models bracket real instrument behavior, and the International Commission on Radiation Units and Measurements (ICRU) describes both in its treatment of particle counting. They differ in one question: does an event that arrives during the dead period extend it? ¹

For facilities running a radiation-safety program, understanding this behavior underlies sound instrument selection and QC, which should be planned alongside choosing the right radiation survey meter and a formal survey-meter calibration program.

Nonparalyzable and paralyzable models

In the nonparalyzable model (also called non-extendable, or ICRU Type I), each recorded event is followed by a fixed dead time $τ$ . Events arriving during $τ$ are lost but do not prolong the dead period. As the true rate climbs, the instrument spends an ever-larger fraction of time dead, and the observed rate rises toward a hard ceiling of $1/ τ$ . ^{1, 2}

In the paralyzable model (also called extendable, or ICRU Type II), every event restarts the dead period, whether or not it was recorded. The detector becomes ready again only after a quiet interval of at least $τ$ . At moderate rates the two models behave similarly, but at very high rates the paralyzable instrument's observed rate reaches a maximum and then decreases, because the detector is almost never given a gap long enough to recover. ^{1, 2}

Key Technical Principles

The count-loss equations

For the nonparalyzable model, the fraction of time the instrument is dead is $m τ$ , so the lost rate is $n m τ$ , and the balance $m = n - nm τ$ rearranges to a simple, invertible correction:

m = \frac{n}{1 + n τ}, n = \frac{m}{1 - m τ} .

For the paralyzable model, the observed rate equals the true rate times the probability that no event occurred in the preceding interval $τ$ , which for Poisson statistics is $e^{- n τ}$ :

m = n e^{- n τ} .

This expression is not one-to-one: it rises, peaks, and falls. Its maximum occurs at $n = 1/ τ$ , giving a maximum observable rate of

m_{m a x} = \frac{1}{e τ} \approx \frac{0.368}{τ} .

Above that true rate, the displayed value drops — the hallmark of a saturating, paralyzable instrument. The $e^{- n τ}$ form follows directly from the Poisson distribution of random decay intervals, and the same treatment underlies dead-time corrections in nuclear-medicine counting and imaging systems. ^{1, 2, 8}

A worked correction

Suppose a count-rate meter with a nonparalyzable dead time of $τ = 5 μ s$ displays $m = 40, 000$ counts per second over a contaminated surface. Then $m τ = 40, 000 \times 5 \times 1 0^{- 6} = 0.20$ , so

n = \frac{40 , 000}{1 - 0.20} = \frac{40 , 000}{0.80} = 50, 000 s^{- 1} .

The instrument is undercounting by 20% — reading 40,000 when 50,000 events are actually occurring each second. At $m τ = 0.5$ the correction factor is 2 (a 50% loss), and as $m τ \to 1$ the correction diverges: the reading is nearly useless. The table below shows how the loss escalates. ²

Observed rate × dead time ( $m τ$ )	True-to-observed ratio ( $n / m$ , nonparalyzable)	Count loss	Interpretation
0.01	1.01	~1%	Negligible; no correction needed
0.10	1.11	~10%	Correction advisable
0.20	1.25	~20%	Correction required
0.50	2.00	~50%	Instrument near its useful limit
0.90	10.0	~90%	Reading unreliable; change instrument or geometry

Comparing the two models

The table below contrasts the two idealized behaviors, using the ICRU terminology. Real survey instruments usually fall between the two, but Geiger-Mueller detectors are closer to paralyzable, and many scalers and timers behave nonparalyzably. ^{1, 2}

Property	Nonparalyzable (Type I, non-extendable)	Paralyzable (Type II, extendable)
Does an unrecorded event extend the dead period?	No	Yes
Observed rate at very high true rate	Approaches ceiling $1/ τ$	Peaks at $1/ (e τ)$ , then falls toward zero
Correction	$n = m / (1 - m τ)$ , invertible	$m = n e^{- n τ}$ , not one-to-one near saturation
Safety concern	Reading plateaus; hazard underestimated at worst by the ceiling	Reading can fold back; intense field can look clean
Typical example	Many scalers, timing electronics	Geiger-Mueller tube

Measuring dead time: the two-source method

Dead time can be measured without a calibrated source strength using the two-source method, which exploits the fact that true rates add but observed rates do not. Counting source 1 alone ( $m_{1}$ ), source 2 alone ( $m_{2}$ ), and both together ( $m_{12}$ ), with background $m_{b}$ , the nonparalyzable dead time follows (to good approximation, neglecting background) from

τ \approx \frac{m _{1} + m _{2} - m _{12}}{2 m _{1} m _{2}} .

The sources are positioned so that neither moves between the individual and combined counts. The measured $τ$ can then correct future readings, provided the instrument is kept within its usable range. ^{1, 2}

Clinical Impact

When dead time changes a radiation-protection decision

The stakes are highest exactly where fields are strongest. A technologist frisking for contamination after a spill, a radiation safety officer mapping the field near a radiopharmaceutical-therapy patient, or a physicist surveying a hot lab can all encounter count rates high enough for dead-time losses to matter. A 20–50% undercount can turn a reportable contamination level into an apparently acceptable one, or make a decontamination effort look successful when residual activity remains. Correcting the reading, or switching to an instrument with a shorter dead time or an ionization chamber, restores the true picture. ^{2, 3}

The paralyzable fold-back is the more insidious failure. Because a Geiger counter can read low — even near zero — in an extremely intense field, an operator who sees a small number and steps closer may be walking into a hazard. Modern survey meters often include an overload or saturation alarm precisely to guard against this, but the operator must know the behavior exists. ^{1, 2}

High-activity imaging and therapy

Dead time is not only a survey-meter issue. The same physics limits quantitative imaging: in Lu-177 radiopharmaceutical-therapy dosimetry, where administered activities reach several gigabecquerels, uncorrected camera dead time biases the measured count rate low and corrupts absorbed-dose estimates. In one phantom-and-patient study using Lu-177 activities up to about 20.75 GBq, applying different dead-time correction methods changed quantitative SPECT values and absorbed-dose estimates by up to roughly 3.5% in a kidney — small when corrected properly, but a systematic error if ignored. The lesson generalizes: whenever activities are high, dead time must be measured and corrected, not assumed negligible. ⁴

The count-rate performance of imaging systems is characterized precisely because of this physics. Standardized performance tests, such as the NEMA measurements for gamma cameras, quantify the observed-versus-true count-rate curve, the maximum count rate, and the count rate at which losses reach 20%, giving a facility the data it needs to keep clinical acquisitions in the usable region. Quality-assurance programs for radioactivity measurement, including the IAEA guidance for nuclear medicine, similarly treat dead-time characterization as a routine part of instrument commissioning and ongoing QC rather than an afterthought. ^{6, 9}

Practical Optimization Tips

1. Match the instrument to the expected field

Use a Geiger-Mueller detector for sensitive low-level work, but switch to a current-mode ionization chamber for high-dose-rate measurements, because an ion chamber integrates charge and does not suffer pulse dead time. Keep a count-rate instrument within the range where $m τ$ is small.

2. Know your instrument's dead time and ceiling

Obtain the dead time from the manufacturer or measure it with the two-source method, and compute the observed-rate ceiling ( $1/ τ$ or $1/ (e τ)$ ). Treat readings approaching that ceiling as unreliable.

3. Recognize and respect saturation

Train staff that a paralyzable instrument can read low in an intense field. Enable and trust overload alarms, and when a reading seems implausibly low near a known source, increase distance or change instruments rather than assuming the field is weak.

4. Correct when it matters

Apply the nonparalyzable correction $n = m / (1 - m τ)$ when $m τ$ is more than a few percent and the result drives a decision. Document the correction. Do not attempt to "correct" a paralyzable reading that may be past its peak — the result is ambiguous.

5. Reduce the rate instead of correcting it

Often the cleanest fix is geometric: back away, collimate, or use a less sensitive detector so the instrument operates in its linear region where no correction is needed.

Common pitfalls to avoid

Trusting a low reading near a strong source. A saturated paralyzable instrument can fold back toward zero.
Correcting past saturation. Near and beyond the paralyzable peak, the observed rate does not uniquely determine the true rate.
Ignoring dead time in high-activity work. Contamination surveys after large-activity administrations and therapy dosimetry both need correction.
Using a pulse-counting instrument for very high dose rates. An ionization chamber is the right tool there.
Assuming the correction is always small. At $m τ = 0.5$ the true rate is double the reading.

Regulatory Considerations

Radiation-protection regulations require surveys to be made with instruments appropriate to the radiation being measured, which implicitly requires the instrument to be operating within a range where its reading is meaningful. Under 10 CFR 20.1501, a licensee must make surveys reasonable to evaluate radiation levels and concentrations, using calibrated and appropriate instruments. An instrument that is saturating or heavily undercounting is not "appropriate" for the field at hand, so recognizing dead-time limits is part of meeting that requirement. ³

The underlying measurement science is codified in consensus standards rather than regulations: ICRU Report 52 on particle counting defines the extendable and non-extendable dead-time models and their correction formulas, NCRP Report No. 58 treats dead time within radioactivity measurement procedures, the ANSI N323 series addresses performance of radiation survey instruments, and NCRP Report No. 112 addresses their calibration for radiation-protection use. These are the references a defensible instrument program cites when it sets usable ranges and correction practices. ^{1, 5, 7}

Agreement States administer equivalent radiation-protection rules. Of the states DRPS serves, Florida, Maryland, Virginia, California, Nevada, Pennsylvania, New York, and New Jersey are NRC Agreement States, while Washington, DC and Delaware are regulated directly by the NRC for byproduct material; X-ray-producing machines are regulated by each state's radiation-control program. Facilities should confirm instrument calibration and survey requirements with the authority having jurisdiction and coordinate them with the radiation safety officer program and radiation safety training. ³

Frequently Asked Questions (FAQs)

What is dead time in a radiation survey meter?

Dead time is the short interval after a detected event during which the instrument cannot register a second event. Each pulse is followed by a recovery period set by the detector physics and the electronics, and any ionizing event arriving during that period is partly or completely lost. The result is that the displayed count rate is lower than the true count rate, and the gap widens as the field intensifies.

What is the difference between paralyzable and nonparalyzable dead time?

In a nonparalyzable model, the instrument is blind for a fixed time after each recorded event and events arriving during that window are simply lost; the observed rate rises toward a ceiling of 1/tau. In a paralyzable model, every event — recorded or not — restarts the dead period, so at very high rates the observed rate peaks and then falls back toward zero. The paralyzable case is more dangerous because a saturated instrument can read low in an intense field.

Why can a Geiger counter read low in a very strong radiation field?

A Geiger-Mueller detector behaves approximately as a paralyzable system. In an extremely intense field, events arrive so fast that each one keeps re-triggering the dead period before recovery, so the tube can latch up and read near zero or drift downward. An operator who trusts the low reading could walk into a hazardous field, which is why instrument selection and knowing the saturation behavior matter.

How do you correct a count rate for dead time?

For a nonparalyzable instrument with dead time tau, the true rate is recovered from the observed rate using n = m divided by the quantity one minus m times tau. The correction is negligible at low rates but grows rapidly as m times tau approaches one. Paralyzable corrections are more complex because the relationship is not one-to-one near saturation.

How is an instrument's dead time measured?

A common approach is the two-source method: count two sources separately and together. Because true rates add but observed rates do not, the discrepancy yields the dead time. The dead time can then be used to correct future measurements, provided the instrument is operated within its usable range.

Does dead time affect dose-rate readings or only count rates?

It affects any rate-based reading driven by individual pulses, including count-rate meters and pulse-mode dose-rate instruments. Current-mode ionization chambers integrate charge rather than counting pulses, so they do not suffer pulse dead time and are preferred for high-dose-rate measurements.

When should a facility consult a medical physicist about survey instruments?

A qualified medical physicist should help select instruments appropriate to the expected fields, establish calibration and quality-control programs, define usable count-rate ranges, and train staff to recognize saturation. This is especially important for programs handling high activities, such as radiopharmaceutical therapy or large sealed sources.

Key Takeaways

Dead time makes counters undercount. The observed rate $m$ is always at or below the true rate $n$ , and the gap grows with the field.
Two models bracket behavior. Nonparalyzable instruments plateau at $1/ τ$ ; paralyzable instruments peak at $1/ (e τ)$ and then fold back.
The correction is simple only for the nonparalyzable case. $n = m / (1 - m τ)$ ; at $m τ = 0.5$ the true rate is double the reading.
Paralyzable fold-back is a safety trap. A Geiger counter can read low in an intense field; never trust an implausibly low reading near a strong source.
High-activity work needs correction. Contamination surveys after large administrations and Lu-177 therapy dosimetry both require dead-time correction.
Often the best fix is geometry. Back away, collimate, or use an ion chamber to keep the instrument in its linear range.

Conclusion

Dead time is a small number — microseconds — with outsized consequences. At the low count rates of routine work it can be ignored, but in exactly the high-field situations where radiation protection matters most, it can halve a reading or, for a paralyzable instrument, invert it. A radiation-safety program that knows each instrument's dead time, its observable-rate ceiling, and its saturation behavior, and that applies the nonparalyzable correction when a decision depends on it, measures what is really there. The alternative — trusting a raw number near saturation — is how a hot area gets called clean. Choosing the right instrument, respecting its limits, and correcting when it counts are what keep a survey honest.

How DRPS Can Help

Diagnostic Radiation Physics Services helps facilities build radiation-detection programs that give trustworthy readings across the full range of fields they encounter. This includes selecting instruments matched to expected count rates and dose rates, establishing calibration and quality-control programs, defining usable ranges and correction practices, and training staff to recognize instrument saturation, delivered through our radiation safety officer, medical physics consulting, and radiation safety training services.

DRPS supports facilities across our service locations, including Florida, Maryland, Virginia, Washington DC, California, Nevada, New York, Pennsylvania, New Jersey, and Delaware.

An instrument you understand protects people. An instrument you trust blindly can mislead you at the worst possible moment.

Related Resources

References

International Commission on Radiation Units and Measurements. ICRU Report 52: Particle Counting in Radioactivity Measurements. Bethesda, MD: ICRU; 1994. icru.org
Knoll GF. Radiation Detection and Measurement. 4th ed. Hoboken, NJ: John Wiley & Sons; 2010. wiley.com
U.S. Nuclear Regulatory Commission. 10 CFR 20.1501: General (surveys and monitoring). ecfr.gov
Desy A, Bouvet GF, Lafrenière N, Zamanian A, Després P, Beauregard JM. Impact of the dead-time correction method on quantitative 177Lu-SPECT (QSPECT) and dosimetry during radiopharmaceutical therapy. EJNMMI Phys. 2022;9(1):54. doi:10.1186/s40658-022-00484-w. doi.org
National Council on Radiation Protection and Measurements. NCRP Report No. 58: A Handbook of Radioactivity Measurements Procedures. 2nd ed. Bethesda, MD: NCRP; 1985. ncrponline.org
International Atomic Energy Agency. Quality Assurance for Radioactivity Measurement in Nuclear Medicine. Technical Reports Series No. 454. Vienna: IAEA; 2006. iaea.org
National Council on Radiation Protection and Measurements. NCRP Report No. 112: Calibration of Survey Instruments Used in Radiation Protection for the Assessment of Ionizing Radiation Fields and Radioactive Surface Contamination. Bethesda, MD: NCRP; 1991. ncrponline.org
Cherry SR, Sorenson JA, Phelps ME. Physics in Nuclear Medicine. 4th ed. Philadelphia: Elsevier Saunders; 2012. elsevier.com
National Electrical Manufacturers Association. NEMA Standards Publication NU 1: Performance Measurements of Gamma Cameras. Rosslyn, VA: NEMA. nema.org