Radioactive Decay and Activity Calculator

Q: How do I find the original activity from a measurement taken today?

Multiply by e+λt rather than e−λt — the positive exponent — which gives a number LARGER than the one you measured. This is the single commonest error in decay work, because both answers are plausible and of the right order of magnitude: for 370 GBq of iridium-192 and thirty days, the forward answer is 279 GBq and the backward answer is 490, and nothing about either number looks wrong on its own. The page prints both at once with FORWARD and BACKWARD on the labels, so the comparison does the checking for you: the backward figure must always be the bigger one.

The activity left after an interval, the activity the interval implies going backward, and the time to reach a target — from one set of inputs, with twenty-two nuclides’ half-lives from NPL and the direction written into every label.

Last updated 7 Oct 2026

The decay arithmetic here is exact; what you put into it is not. A half-life, an activity quoted for a stated date and the assay time itself all carry uncertainty, and an activity projected far from the calibration date compounds it. Nothing here tells you whether a source is safe to handle or to move, and nothing here replaces the certificate the source came with.

Radioactive decay and remaining activity

Nuclide, activity and interval → all three directions

NuclideTwenty-two nuclides, with their half-lives from NPL Report IR 6. Picking one fills the half-life field below and locks it; choosing the last option unlocks that field so any nuclide at all can be used. The half-lives carry published uncertainties — NPL’s own, which run from 0.0096 per cent on manganese-54 to 0.44 per cent on radium-223, a factor of forty-five between the best and the worst known in this list — and the sources do not fully agree in the last figures. The default here is iridium-192, because its seventy-four-day half-life is the one that costs money: an industrial radiography source loses a quarter of its output in a month and is replaced two or three times a year, and the replacement schedule IS this calculation.

Half-lifeLocked while a nuclide is selected above. Choose “other nuclide” and this becomes the input, which is how an unlisted nuclide — a short-lived cyclotron product, a daughter, anything at all — still works here. The decay arithmetic does not care where the half-life came from.

…inHalf-lives here are stored in days, which is how NPL tabulates them. A year means 365.25 days, the Julian year; with 365 instead, a fifty-year projection of a thirty-year half-life moves by about a third of a per cent.

The activity you knowRead this as whichever end you have. The page computes the activity one interval LATER and the activity one interval EARLIER from the same number, and labels both, so it does not matter whether your figure is a certificate value or a measurement taken today.

…inOne curie is 3.7×10¹⁰ becquerel by definition, not by measurement, so the curie columns on this page are a rename of the becquerel ones and carry no extra uncertainty of their own.

IntervalA number and a unit, not a date: the engine behind this page has no date type and cannot subtract one timestamp from another. For the hours-and-minutes case that nuclear medicine actually lives in, the decay-correction page on this site takes the two intervals separately and keeps their directions apart.

…inA year here is 365.25 days. That matters only on long projections, and then only in the third figure: over fifty years of caesium-137, using 365 instead moves the answer by about 0.1 per cent.

Target activity, for the third answerIn the same unit as the activity above. The page returns how long it takes to fall from the one to the other — and returns a NEGATIVE interval if the target is higher, because then the question is how long ago it was that high.

Not a circuit: the decay curve of your own nuclide, drawn to scale over ten half-lives, with a live crosshair at the interval you entered. The reason to draw it rather than tabulate it is that the shape is the argument. Half of the activity is gone in the first tenth of the x axis. The dashed lines are the successive halvings, and notice how fast they crowd together at the bottom: getting from 100 per cent to 50 costs one half-life, and getting from 6.25 per cent down to 0.098 costs six more. That is why waiting is a cheap disposal route for a six-hour nuclide and a useless one for a thirty-year nuclide, and it is why the curve never reaches the axis at all. The ten-half-lives mark is the figure everyone quotes for decay in storage; it leaves 0.098 per cent, one part in 1,024, and it is no longer a regulatory requirement. The figures underneath are your own, including both directions: forward is smaller than the number you entered and backward is larger, every time.

279.2GBqExample

370 GBq of iridium-192 — a typical industrial radiography source — thirty days after its certificate date, with 100 GBq set as the target

One exponential, two constants, and the three ways out of it

A = A₀·e^−λt · λ = ln2 / t½ · A / A₀ = 2⁻ⁿ, where n = t / t½

A₀, A: the activity at the two ends of the interval, in any unit you like as long as it is the same unit at both ends. The ratio is what the physics computes; the unit cancels. One curie is 3.7×10¹⁰ becquerel by definition, so the curie columns on this page are a rename and not a second measurement
λ: the decay constant, the probability per unit time that any one atom decays. It is a property of the nuclide and of nothing else: not of temperature, not of pressure, not of chemical form, and not of how much you have. That is the whole reason this arithmetic is exact
t½: the half-life. ln2/λ, and therefore 0.6931 times the mean life rather than equal to it — the two get confused, and the mean life is the longer of them by 44 per cent
n = t/t½: the number of half-lives elapsed, and the only number on this page worth memorising. It makes the answer a power of a half: n = 1 leaves 0.5, n = 10 leaves 0.00098, n = 20 leaves a millionth. Everything else here is putting a clock on that
e^(+λt): the backward direction, and the sign error that this page is laid out to prevent. Projecting back to a calibration date needs the POSITIVE exponent and gives a number LARGER than the one you entered. If your answer came out smaller going backward, the sign is wrong

Worked example

370 GBq of iridium-192 — a typical industrial radiography source — thirty days after its certificate date, with 100 GBq set as the target
THE HALF-LIFE. NPL gives iridium-192 as 73.827 ± 0.013 days. So λ = ln2/73.827 = 0.00938880 per day, and the mean life is 73.827/ln2 = 106.510 days — 44 per cent longer than the half-life, which is the usual place that pair gets mixed up
THE INTERVAL IN HALF-LIVES, which is the number that makes the rest intuitive. Thirty days is 30/73.827 = 0.4064 half-lives. Not quite half a half-life, so expect a bit more than 70 per cent to survive
THE SURVIVING FRACTION. Two routes, and they must agree. e^−λt = exp(−0.00938880×30) = 0.754527; and 2^−0.4064 = 0.754527. They are the same number written two ways, and checking a decay calculation against the power of a half is the cheapest sanity test there is
FORWARD. 370 GBq × 0.754527 = 279.2 GBq. A quarter of the source's output has gone in a month. For a radiography crew that is not an abstraction — the exposure time for the same film density has gone up by 1/0.7545 = 1.325 times
BACKWARD, and this is the direction to be careful about. If the 370 GBq is what you measured TODAY rather than what the certificate said, then thirty days ago the source held 370 GBq × e^+λt = 490.4 GBq. Larger, by the same factor the forward answer was smaller by. Both numbers are on the page above, labelled FORWARD and BACKWARD, precisely because 279.2 GBq and 490.4 GBq are both perfectly plausible answers and nothing but the label tells them apart
TO THE TARGET. Falling from 370 to 100 GBq is a factor of 3.7, so t = ln(3.7)/λ = 139.35 days, which is 1.888 half-lives or 0.382 years. Sources are usually exchanged on a schedule rather than on a number, and this is the number behind the schedule
WHAT THE UNCERTAINTY IS WORTH. NPL's ±0.013 days on the half-life is 0.018 per cent. Pushed through a thirty-day projection it moves the answer by about 0.01 per cent — four orders of magnitude below the uncertainty on the source certificate itself. The decay arithmetic is never the weak link over an interval of this length. Over six hundred years of caesium-137 it is a different story, and the half-life page carries that case

What is left after n half-lives, and how long that is

The first four columns are pure arithmetic and are the same for every nuclide that exists; the last three put a clock on them for three nuclides whose half-lives are six hours, seventy-four days and thirty years. Two rows are worth stopping at. TEN HALF-LIVES leaves 0.098 per cent, one part in 1,024, and it is the number everybody quotes for decay-in-storage — but it is no longer a rule. 10 CFR 35.92 was read in full for this page and contains no holding period whatever: it allows material with a half-life of 120 days or less to be held “without regard to its radioactivity” provided the licensee monitors it at the surface and finds that its radioactivity “cannot be distinguished from the background radiation level”. NRC Regulatory Issue Summary 2004-17 Rev 1 says plainly what happened: the rulemaking “eliminat[ed] the requirement to hold radioactive waste for a period of 10 half lives before disposal” and the “License Condition 140 stipulation that decay-in-storage waste be held for 10 half-lives will be deleted”. So ten half-lives is a good rule of thumb and a survey meter is the actual test. The other row is the LAST one: twenty half-lives leaves a millionth, and no row anywhere on this table reaches zero. The decay arithmetic is exact. The inputs are not: a quoted activity carries the calibration laboratory’s uncertainty, the assay time is often recorded only to the nearest hour, and a half-life itself has a published uncertainty. Projected far from the calibration date those compound — and the published half-lives themselves disagree in the last figures, caesium-137 being 10976 days to NPL and 11000 to another published decay sheet.
Half-lives	Fraction left	Per cent left	One part in	…for Tc-99m	…for Ir-192	…for Cs-137
0	1	100	1 in 1.0	0.00 h	0.00 d	0.0 y
1	0.5	50	1 in 2.0	6.01 h	73.83 d	30.1 y
2	0.25	25	1 in 4.0	12.01 h	147.65 d	60.1 y
3	0.125	12.5	1 in 8.0	18.02 h	221.48 d	90.2 y
4	0.0625	6.25	1 in 16.0	24.03 h	295.31 d	120.2 y
5	0.03125	3.125	1 in 32.0	30.03 h	369.13 d	150.3 y
6	0.015625	1.5625	1 in 64.0	36.04 h	442.96 d	180.3 y
6.6439 (the 1 per cent point)	0.01	1	1 in 100.0	39.91 h	490.50 d	199.7 y
7	0.0078125	0.78125	1 in 128.0	42.05 h	516.79 d	210.4 y
8	0.00390625	0.390625	1 in 256.0	48.05 h	590.62 d	240.4 y
9	0.001953125	0.195313	1 in 512.0	54.06 h	664.44 d	270.5 y
9.9658 (the 0.1 per cent point)	0.001	0.1	1 in 1,000.0	59.86 h	735.74 d	299.5 y
10 (the old ten-half-lives rule)	0.0009765625	0.097656	1 in 1,024.0	60.07 h	738.27 d	300.5 y
12	0.0002441406	0.024414	1 in 4,096.0	72.08 h	885.92 d	360.6 y
15	0.0000305176	0.003052	1 in 32,768.0	90.10 h	1,107.40 d	450.8 y
19.9316 (the one-millionth point)	0.000001	0.0001	1 in 1,000,000.0	119.72 h	1,471.49 d	599.0 y
20	0.0000009537	0.000095	1 in 1,048,576.0	120.13 h	1,476.54 d	601.0 y

Twenty-two nuclides: how much is left after a day, a week, a month and a year

Read down the four percentage columns and the practical consequence of a half-life becomes obvious in a way that the half-life column alone does not convey. Fluorine-18 is down to 0.0112 per cent in a single day and to nothing at all in a week, which is why a PET centre needs a cyclotron within a few hours’ drive and why the dose is ordered rather than stocked. Technetium-99m is 6.27 per cent after a day — four half-lives, near enough — which is why a generator is eluted every morning rather than every week. Iridium-192 is 93.6 per cent after a week and 75.5 per cent after a month, which is a radiography crew’s exposure-time chart being rewritten every few weeks. Caesium-137 is 97.7 per cent after a year, which is why it is the reference source. And americium-241 loses 0.16 per cent in a year, which is why the one in a smoke detector is still working. The decay-mode column is there because it decides what the activity MEANS outside the container: two of these nuclides, yttrium-90 and strontium-90, emit no useful gamma at all. The decay arithmetic is exact. The inputs are not: a quoted activity carries the calibration laboratory’s uncertainty, the assay time is often recorded only to the nearest hour, and a half-life itself has a published uncertainty. Projected far from the calibration date those compound — and the published half-lives themselves disagree in the last figures, caesium-137 being 10976 days to NPL and 11000 to another published decay sheet.
Nuclide	Half-life	…in days	Decay constant, per day	% left after 1 day	after 7 days	after 30 days	after 1 year	Decay mode
Tc-99m	6.0067 h	0.25028	2.76948690	6.2694	0.0000	0.0000	0.0000	IT
F-18	1.8288 h	0.07620	9.09641969	0.0112	0.0000	0.0000	0.0000	B+
Ga-67	3.2613 d	3.26130	0.21253708	80.8530	22.5878	0.1702	0.0000	EC
In-111	2.8049 d	2.80490	0.24712010	78.1047	17.7313	0.0603	0.0000	EC
Tl-201	3.0421 d	3.04210	0.22785154	79.6242	20.2916	0.1075	0.0000	EC
Mo-99	2.7479 d	2.74790	0.25224614	77.7053	17.1063	0.0517	0.0000	B-
Y-90	2.6684 d	2.66840	0.25976135	77.1236	16.2297	0.0413	0.0000	B-
Lu-177	6.6470 d	6.64700	0.10427970	90.0973	48.1929	4.3788	0.0000	B-
I-131	8.0223 d	8.02230	0.08640255	91.7225	54.6174	7.4864	0.0000	B-
Ra-223	11.4300 d	11.43000	0.06064280	94.1159	65.4097	16.2142	0.0000	A
I-125	59.4070 d	59.40700	0.01166777	98.8400	92.1572	70.4665	1.4099	EC
Ir-192	73.8270 d	73.82700	0.00938880	99.0655	93.6391	75.4527	3.2411	B-
Co-57	271.8000 d	271.80000	0.00255021	99.7453	98.2307	92.6347	39.3976	EC
Mn-54	312.1300 d	312.13000	0.00222070	99.7782	98.4575	93.5550	44.4364	EC
Zn-65	244.0100 d	244.01000	0.00284065	99.7163	98.0312	91.8311	35.4323	EC
Na-22	2.6026 y	950.60000	0.00072917	99.9271	99.4909	97.8362	76.6187	B+
Co-60	5.2709 y	1,925.20000	0.00036004	99.9640	99.7483	98.9257	87.6776	B-
Ba-133	10.5399 y	3,849.70000	0.00018005	99.9820	99.8740	99.4613	93.6352	EC
Eu-152	13.5222 y	4,939.00000	0.00014034	99.9860	99.9018	99.5799	95.0032	EC
Sr-90	28.8022 y	10,520.00000	0.00006589	99.9934	99.9539	99.8025	97.6221	B-
Cs-137	30.0507 y	10,976.00000	0.00006315	99.9937	99.9558	99.8107	97.7198	B-
Am-241	432.5804 y	158,000.00000	0.00000439	99.9996	99.9969	99.9868	99.8399	A

The same equation, rearranged three ways

One equation, two unknowns at a time, and five questions. The rows are ordered by how often they are asked and the second one is the dangerous one: e^+λt rather than e^−λt, and both are plausible-looking numbers of the right order of magnitude. That is exactly why this page prints both at once with the direction in capitals at the start of each label — if the two answers are on screen together, the one that is bigger is the one that looks backward in time, and no arithmetic is needed to tell them apart.
What you want	Rearrangement	On this page
The activity after an interval	A = A₀·e^−λt	the headline, and the FORWARD rows
The activity before an interval — the starting figure a present measurement implies	A₀ = A·e^+λt	the BACKWARD rows. Note the sign: this is the one that is LARGER, and getting it the wrong way round is the commonest error in the whole subject
The interval to a target activity	t = ln(A₀/A) / λ = log₂(A₀/A)·t½	the TO THE TARGET rows, in four units including half-lives
The half-life, from two activities and the interval between them	t½ = t·ln2 / ln(A₀/A)	not here — it is the measurement rather than the prediction, and it belongs on the half-life page
The number of atoms behind an activity	N = A / λ	not here — the specific-activity page, which also turns it into a mass

Three directions out of one exponential, and the one number worth carrying in your head

This page answers the decay question in all three directions at once, and that is a deliberate choice rather than a shortcut. A = A₀e^−λt has three quantities in it and a reader arrives holding any two. A mode selector would make them choose before they have read anything; printing all three answers from one set of inputs means nobody has to. More usefully, it puts the forward and the backward answer side by side — and since the only difference between them is the sign of the exponent, and both are plausible numbers of the right order, having them on screen together with the direction in capitals is the cheapest protection there is against the commonest mistake in the subject.

Count half-lives, not days. The one number worth carrying around is n = t/t½, the number of half-lives elapsed, because it turns the answer into a power of a half that can be done in the head: one half-life leaves 50 per cent, two leave 25, three leave 12.5, and ten leave 0.098. Every exponential on this page is that, with a clock attached. It is also the sanity check: if the calculator says something and 2⁻ⁿ says something else, one of the two inputs is in the wrong unit. Both numbers are printed above for exactly that comparison.

Ten half-lives is no longer the rule, and this is worth knowing if you work anywhere near a decay-in-storage cupboard. The ten-half-lives holding period is quoted everywhere as though it were regulation. It was: it was License Condition 140 in the United States. It is not any more. 10 CFR 35.92 as it now stands sets no holding period at all — material with a half-life of 120 days or less may be held “without regard to its radioactivity” so long as the licensee monitors it at the surface and determines that its radioactivity “cannot be distinguished from the background radiation level” with a survey meter on its most sensitive scale and no shielding in the way, and removes the labels. NRC Regulatory Issue Summary 2004-17 Revision 1 describes the change as “eliminating the requirement to hold radioactive waste for a period of 10 half lives before disposal”. Ten half-lives remains an excellent first guess at how long the wait will be. The test at the end of it is an instrument.

The arithmetic is exact and the inputs are not. Radioactive decay is the best-behaved process in applied physics: λ is a property of the nuclide and of nothing else, so the answer does not depend on temperature, pressure, chemical form or how much you have. What does carry error is everything around it. A certificate activity has the calibration laboratory’s uncertainty on it, typically a few per cent. An assay time is often recorded to the nearest hour, which for technetium-99m is a sixth of a half-life and therefore eleven per cent of the activity. And the half-life itself has a published uncertainty, which is multiplied by the number of half-lives before it reaches the answer — small over a month, not small over a century. The page prints NPL’s uncertainty and what it is worth on your particular interval, so that last term can be read rather than assumed.

What this page cannot do, said plainly. It cannot take a date: the engine has no date type, so an interval here is a number and a unit, and the hours-and-minutes case that nuclear medicine lives in belongs on the decay-correction page instead. It models ONE nuclide decaying on its own, so it is wrong wherever a daughter matters — a molybdenum-99 generator is not a molybdenum-99 source, strontium-90 comes with yttrium-90 in equilibrium, and caesium-137’s famous 662 keV photon is emitted by barium-137m and not by caesium at all. And it says nothing about what the activity will do to anybody: activity is not dose, and the step from one to the other needs the emissions, the geometry and the distance, which are other pages’ subjects.

Frequently asked questions

How do I work out the activity remaining after a given time?

Divide the elapsed time by the half-life to get n, the number of half-lives, then multiply the starting activity by 2 raised to the power of minus n — which is the same thing as A₀e^−λt with λ = ln2/t½. For 370 GBq of iridium-192 after thirty days: n = 30/73.827 = 0.4064, 2^−0.4064 = 0.7545, and 370 × 0.7545 = 279 GBq. The two things to check are that the time and the half-life are in the SAME unit, and that you have not used the mean life by mistake — it is 44 per cent longer than the half-life and using it gives an answer that is wrong in the right direction to look believable.

What is the difference between the half-life and the mean life?

The half-life is when half the atoms have gone; the mean life is the average lifetime of an atom, and it is 1/λ = t½/ln2 = 1.4427 t½. So the mean life is always the longer of the two, by 44.27 per cent, for every nuclide that exists — the ratio does not depend on the nuclide at all. The reason the mean life exists is that it is what appears naturally in the exponent: the activity falls to 1/e, about 36.8 per cent, in one mean life. Substituting one for the other in an exponential is a 44 per cent error in the exponent.

Can I put in a calibration date and a use date instead of an interval?

Not on this page, and the reason is worth stating rather than working around: the calculation engine behind this site is numeric only and has no date type, so it cannot subtract one timestamp from another. You give it the interval. For the case where that is genuinely awkward — a vial assayed at 07:00 for a patient at 14:20, where the two intervals point in opposite directions and the sign is the thing people get wrong — the decay-correction page on this site takes assay-to-now and now-to-administration separately and labels each direction explicitly.

Does radioactive decay speed up if the material is hot, or compressed, or chemically bound?

No, and that is what makes this arithmetic exact rather than approximate. The decay constant is a property of the nucleus. Chemistry happens in the electron cloud, and temperature and pressure act on the material rather than the nucleus, so none of them reaches λ. There are real exceptions and they are tiny and exotic: electron-capture nuclides depend very slightly on electron density, so a fully ionised atom cannot electron-capture at all, and a few internal-conversion rates shift by small fractions of a per cent with chemical environment. Nothing you can do in a laboratory changes a half-life usefully.

Is ten half-lives still the rule for decay in storage?

It is a good rule of thumb and it is no longer a requirement in the United States. Ten half-lives leaves 0.098 per cent, one part in 1,024, which is why the figure became the convention. But 10 CFR 35.92 as it now stands has no holding period in it: material with a half-life of 120 days or less may be held “without regard to its radioactivity” provided the licensee surveys it at the surface and finds it indistinguishable from background on a meter’s most sensitive scale with no shielding, and removes the labels. NRC Regulatory Issue Summary 2004-17 Rev 1 is explicit that the rulemaking eliminated the ten-half-lives requirement and deleted the licence condition that carried it. Your own licence and local rules may still impose it, and they govern.

How do I find the original activity from a measurement taken today?

Multiply by e^+λt rather than e^−λt — the positive exponent — which gives a number LARGER than the one you measured. This is the single commonest error in decay work, because both answers are plausible and of the right order of magnitude: for 370 GBq of iridium-192 and thirty days, the forward answer is 279 GBq and the backward answer is 490, and nothing about either number looks wrong on its own. The page prints both at once with FORWARD and BACKWARD on the labels, so the comparison does the checking for you: the backward figure must always be the bigger one.

Why does the activity never reach zero?

Because the model is a constant probability per atom per unit time, which gives an exponential, and an exponential has no zero. Physically, of course, the last atom does eventually decay: the exponential is a description of a very large number of atoms and stops being meaningful when only a few are left. The useful consequence is practical rather than philosophical. Waiting is an excellent way to dispose of a six-hour nuclide and a useless one for a thirty-year nuclide, and no number of half-lives turns the second into the first — which is why the test for decay in storage is now an instrument reading against background rather than a number of half-lives on a calendar.

Does this work for a generator, or for a parent and daughter together?

No. This page models one nuclide decaying by itself. A molybdenum-99/technetium-99m generator is the standard counter-example: the technetium activity in it is set by ingrowth from the molybdenum as well as by its own decay, it reaches a maximum about 22.8 hours after each elution rather than falling monotonically, and transient equilibrium is reached at about 48 hours. Only about 87 per cent of molybdenum-99 decays go to the metastable technetium at all. Strontium-90 is the same problem with a long clock on it — its yttrium-90 daughter sits in equilibrium with it — and even caesium-137’s famous 662 keV line is emitted by barium-137m rather than by caesium. For any of those, a two-nuclide treatment is needed and this page will mislead.

Related calculators

References

A. Pearce, NPL Report IR 6: Recommended Nuclear Decay Data, National Physical Laboratory. Cited, not reproduced — Crown copyright. It is the source of every half-life on this page and, more usefully, of every half-life’s UNCERTAINTY, which most published tables drop. Those uncertainties are the reason the nuclide table here prints a ± column: manganese-54 is known to 0.03 days in 312.13, which is 0.0096 per cent, and technetium-99m to 0.000 04 days in 0.250 28, which is 0.016; caesium-137 is known to 30 days in 10 976, which is 0.27 per cent, and radium-223 to 0.05 days in 11.43, which is 0.44 and is the widest in the set — a factor of forty-five across the list, and a projection over many half-lives inherits whichever end it lands on.
10 CFR 35.92, Decay-in-storage (US Nuclear Regulatory Commission; a US Government work, quoted). It permits a licensee to hold byproduct material “with a physical half-life of less than or equal to 120 days” for decay in storage “without regard to its radioactivity” provided the licensee “monitors byproduct material at the surface before disposal and determines that its radioactivity cannot be distinguished from the background radiation level with an appropriate radiation detection survey meter set on its most sensitive scale and with no interposed shielding”, and removes or obliterates the radiation labels. Read in full for this batch because of what is NOT in it: there is no ten-half-lives holding period anywhere in the section.
NRC Regulatory Issue Summary 2004-17, Revision 1, Revised Decay-in-Storage Provisions for the Storage of Radioactive Waste Containing Byproduct Material, Office of Nuclear Material Safety and Safeguards, 27 September 2005 (a US Government work, quoted). This is the document that settles the ten-half-lives question. It describes the rulemaking as “eliminating the requirement to hold radioactive waste for a period of 10 half lives before disposal”, replaced by the rule that “the waste must be held in storage until the radiation exposure rate cannot be distinguished from background radiation levels”, and states that the “License Condition 140 stipulation that decay-in-storage waste be held for 10 half-lives will be deleted”. Ten half-lives was therefore real, and is now history: the test is a survey meter, not a calendar.
G. J. Morrissey, “Lest We Forget Generator Technology”, Journal of Nuclear Medicine Technology, September 1996. The source for the two timings this batch quotes about a molybdenum-99/technetium-99m generator: “the maximum amount of activity is achieved at approximately 23 hr after the previous elution” — 22.83 hours in its own table — and “transient equilibrium is attained with the ⁹⁹Mo/^99mTc generator system at approximately 48 hr following the last elution”, at which point “the ^99mTc activity approximates the ⁹⁹Mo activity”. It also makes the point that the technetium mole fraction of the eluate is about 70 per cent after a four-hour ingrowth and about 27 per cent after twenty-four, which is why the chemistry and the activity do not keep step.
Technetium-99m, Wikipedia (read 7 October 2026). Read for the generator description and for the branching figure a single-nuclide decay calculation silently assumes away: “over 87% of the decays lead to the desired ^99mTc”. It also gives the half-life as 6.0066 hours, against NPL’s 6.0067, and notes that a generator “must be replaced weekly, since the half-life of ⁹⁹Mo is still only 66 hours”.
Yttrium-90, Wikipedia (read 7 October 2026). Read for the one thing a decay page has to get right about yttrium-90: it is not a gamma emitter. “Although it decays to the 1.7 MeV excited 0+ state of ⁹⁰Zr with frequency more than 0.01%, emission of a gamma ray is forbidden”, and “the useful photons emitted through this isotope’s decay are instead bremsstrahlung X-rays”. That is why the decay arithmetic here applies to it unchanged while a gamma dose-rate calculation does not apply to it at all.
The 2019 revision of the SI, as published by the BIPM and tabulated in NIST’s Fundamental Physical Constants. Two exact values are used here and both are exact by definition rather than by measurement: the Avogadro constant N_A = 6.022 140 76×10²³ mol⁻¹, and the elementary charge, through which 1 eV = 1.602 176 634×10⁻¹⁹ J. One curie is likewise 3.7×10¹⁰ Bq by definition, so every curie figure on these pages is a rename of a becquerel figure and adds no uncertainty of its own.