Bearing Relubrication Interval and Grease Quantity Calculator

The grease interval from the published closed form, halved for every 15 °C above 70, with the dose G = 0.005·D·B, the bearing’s own free space from its catalogue mass, and the viscosity ratio κ the a_ISO chart needs.

Last updated 24 Sep 2026

Bearing relubrication interval and grease quantity

Size, speed, temperature → interval and dose

Bearing bore d (mm)The interval formula uses the BORE, not the mean diameter — the speed factor n·d_m is what the chart form uses and it is printed below for comparison.

Outside diameter D (mm)

Width B (mm)

Bearing mass (g)From the catalogue row. It is the only honest way to get the bearing’s free internal volume: the annulus it occupies, less the steel actually in it.

Speed (rev/min)

Bearing typeA ball bearing tolerates grease best and gets the longest interval; a tapered or spherical roller bearing has more sliding at the roller ends and gets a tenth of it. NOTE that the two published sources disagree on the last one: STLE’s K values are 10, 5 and 1, and SKF’s chart type factors are in the ratio 1, 0.5 and 0.2 — the same for the first two and a factor of two apart on the third.

Operating temperature at the outer ring (°C)The chart and the formula are referred to 70 °C. Above that the interval HALVES for every 15 °C, which is a brutal rule: 100 °C is a quarter of the interval and 130 °C is a sixteenth.

EnvironmentThe contamination and vibration correction. These are representative values for the published correction factor f_C rather than a standard’s table — a wet or dusty bearing needs relubricating far more often because the grease’s job includes flushing contaminant out of the bearing.

Oil or grease base oil viscosity at operating temperature ν (mm²/s)The ACTUAL kinematic viscosity at the operating temperature, not the ISO VG grade number (which is the viscosity at 40 °C). Viscosity unit conversion belongs to the converters plugin and is not linked from here.

Rated viscosity ν₁ from the bearing maker’s chart (mm²/s)READ FROM A CHART, against the mean diameter and the speed. This page will not invent it, for the same reason it will not invent a_ISO. The default here is deliberately equal to ν so that κ = 1, which is the boundary of adequate lubrication — replace it with your own figure.

Not a circuit: two number lines. The top one is the speed factor n·d_m with the three published bands marked — below 200,000 an interval of a year or more, 200,000 to 400,000 five to twelve months, above 400,000 weeks — and a pointer at where your bearing sits. It is worth computing that before anything else, because it tells you whether you have a sealed-for-life bearing, a maintenance schedule or an automatic lubricator, before any formula is involved. Below it is the temperature rule drawn as a ladder: the interval halves for every 15 °C the outer ring runs above 70, so each rung is half the one above and the bar beside it is half as long. Every figure on that ladder is your own bearing's interval at that temperature, computed live. Read the bottom rung against the top one: thirty degrees of temperature costs three quarters of the interval, and sixty degrees costs fifteen sixteenths of it. Nothing else you can change to a greased bearing has that much leverage.

13,666hExample

A 40 × 80 × 18 mm ball bearing weighing 370 g, at 1,450 rev/min, outer ring at 70 °C, clean and dry

The interval, the dose, and the lubrication condition

t_f = K[(14×10⁶/(n√d)) − 4d] · t_adj = t_f·f_T·f_C · f_T = 2^(−(T−70)/15) · n·d_m with d_m = (d+D)/2 · G = 0.005·D·B grams · κ = ν/ν₁

t_f: base relubrication interval in hours, at 70 °C in clean conditions. It can come out NEGATIVE, and that is the formula telling you the speed factor has passed the point where grease alone will do
K: the bearing type factor: 10 for ball bearings, 5 for cylindrical and needle roller, 1 for tapered and spherical roller. More sliding contact means a shorter interval
d: the BORE, in millimetres, in the interval formula. The speed factor uses the MEAN diameter instead, which is a different number and a different purpose
f_T: halved for every 15 °C above 70 °C, and never above 1. The published factors 1, 0.5, 0.25 and 0.125 are exactly this expression
G = 0.005·D·B: the dose in grams, with D the OUTSIDE diameter and B the width, both in millimetres. Not the bore: one published description says bore and is out by a factor of two
κ: viscosity ratio. The actual viscosity at temperature over the rated viscosity from the maker’s chart. One of the two arguments a_ISO is read against

Worked example

A 40 × 80 × 18 mm ball bearing weighing 370 g, at 1,450 rev/min, outer ring at 70 °C, clean and dry
Compute the speed factor first, because it tells you what kind of problem you have. d_m = (40+80)/2 = 60 mm, so n·d_m = 1,450 × 60 = 87,000 mm/min. That is below 200,000, which puts this bearing in the band where the published guidance is an interval of a year or more and where grease-for-life is realistic
Now the interval itself: t_f = K[(14×10⁶/(n√d)) − 4d] with K = 10 for a ball bearing. 14,000,000/(1,450 × √40) = 1,526.6, less 4 × 40 = 160, gives 1,366.6, times 10 = 13,666 hours. About 569 days, or 1.56 years of continuous running — which agrees with what the speed factor band predicted
Note the shape of that expression, because one published rendering of it moves a bracket. Written as above, the interval goes to ZERO at a finite speed — for a 40 mm bore, at 13,835 rev/min — and that is the formula telling you something true: past that speed factor there is no grease interval at all and you need oil. The mis-bracketed version, (14×10⁶ − 4d)/(n√d), can never go negative and so never says it
Corrections. At 70 °C the temperature factor is exactly 1, because 70 is the reference. Raise the outer ring to 85 °C and it is 0.5; to 100 °C and it is 0.25. Clean and dry gives an environment factor of 1. So the adjusted interval here is the same 13,666 hours — but move this bearing into a dusty, vibrating place at 100 °C and it becomes 1,708 hours, which is 71 days
THE DOSE: G = 0.005·D·B = 0.005 × 80 × 18 = 7.2 grams, using the OUTSIDE diameter. Read that formula with the bore instead and you get 3.6 grams, exactly half, and one published description of the formula does say bore
How much the bearing can hold, for comparison: the annulus it occupies is π/4(80² − 40²) × 18 = 67.9 cm³, and 370 g of steel at 7.8 g/cm³ is 47.4 cm³ of that, leaving 20.4 cm³ free. That is an upper bound — the cage takes some of it and grease cannot be packed right up to a seal — but it sets the scale: the housing around it is filled to 30 to 50 per cent of ITS free space, not to 100
AND THE REASON THE QUANTITY MATTERS AS MUCH AS THE FREQUENCY. Grease that cannot escape is churned by the rolling elements; churning generates heat; heat oxidises the base oil and bleeds it out of the thickener; what is left is a hard soap. A bearing packed solid runs hotter than one filled correctly and fails sooner. Over-greasing is one of the commonest causes of bearing failure, and it is done with the best of intentions
Finally the lubrication condition, which links to the life calculation. κ = ν/ν₁ with both at the operating temperature. The defaults here give κ = 1.00, the boundary of a full film — but ν₁ has to come from the bearing maker's chart and this page will not invent it, for the same reason the life page will not invent a_ISO. κ is one of that chart's two arguments, so the two pages meet here

The temperature rule: halved for every 15 °C above 70 °C

This is the most violent factor on the page and the one most often left out. SKF’s rule is a halving for every 15 °C above 70; the published factors 1.0, 0.5, 0.25 and 0.125 at 70, 85, 100 and 115 °C are exactly 2^(−ΔT/15), which is how this page computes it. Note one disagreement in the sources: where SKF gives a single 15 °C, Fabrico’s summary of the same material says grease life halves for every 10 to 15 °C, and the 10 °C version is twice as harsh at 30 °C above the reference. Both are published; this page uses 15 and says so. The physical reason is that grease life is governed by oxidation of the base oil and by the thickener losing its ability to hold it, and both are chemical rates — which is why the rule is a halving per fixed temperature step rather than anything linear. It also means the temperature to use is the one at the OUTER RING, not the ambient and not the housing surface. A drive is rated for the duty it sees, not for the power it nominally transmits. The service factor used here is stated; the manufacturer’s own factor for your machine class and daily running hours takes precedence over any general table.
Outer ring temperature	Factor f_T	Interval for this page’s example	In days	As % of the 70 °C interval
50 °C	1.0000	13,666 h	569.4 d	100.0%
70 °C	1.0000	13,666 h	569.4 d	100.0%
85 °C	0.5000	6,833 h	284.7 d	50.0%
100 °C	0.2500	3,417 h	142.4 d	25.0%
115 °C	0.1250	1,708 h	71.2 d	12.5%
130 °C	0.0625	854 h	35.6 d	6.3%
145 °C	0.0313	427 h	17.8 d	3.1%

What the speed factor n·d_m tells you before you calculate anything

d_m = (d+D)/2 and n is in rev/min, so the speed factor is a surface-speed proxy with the units of mm/min. It is worth computing first, because it tells you which kind of problem you have before any formula is involved. Below 200,000 the honest answer is usually a sealed, greased-for-life bearing and no maintenance schedule at all. Between 200,000 and 400,000 you need a relubrication plan, and the quantity matters as much as the frequency because over-greasing is itself a failure mode. Above 400,000 grease is at its limit and the question becomes oil — circulating, jet or mist — which is a different design. This page’s example, a 40 × 80 mm bearing at 1,450 rev/min, gives 87,000: comfortably in the first band. These dimensions come from a published standard’s table, not from a formula. The standard itself is cited below and the printed values are attributed to the catalogue they were taken from; a different publisher may round differently in the last digit.
n·d_m (mm/min)	Interval, order of magnitude	What that means in practice
up to 200,000	a year or more	Grease for life is realistic; a sealed bearing will outlast its grease only marginally
up to 400,000	five to twelve months	Relubrication has to be planned and the quantity matters
above 400,000	weeks	Continuous or very frequent feed, or oil. A grease nipple and a maintenance round will not keep up

Grease quantity: the dose, and the trap in the formula

The dose formula is G = 0.005·D·B grams with D and B in millimetres, and D is the bearing’s OUTSIDE diameter. Worth labouring, because this batch found a published description of the same formula that says to “multiply the bore diameter by the bearing width and a constant”, which on the bearings in this table gives between a half and a third of the right answer. The continuous-feed column is a different quantity again — (0.3 to 0.5)·D·B×10⁻⁴ grams per HOUR, for a bearing on an automatic lubricator or a central system — and it is not the dose divided by the interval. AND THE MORE IMPORTANT POINT: over-greasing is one of the commonest causes of bearing failure. Grease that cannot escape gets churned by the rolling elements, churning generates heat, heat oxidises the base oil and bleeds it out of the thickener, and what is left is a hard soap that no longer lubricates. A bearing packed solid runs hotter than one filled correctly, which is why the housing is filled to 30 to 50 per cent of its free space and not to 100 — and why a relief path for the old grease matters as much as the nipple does. This page sizes a part; it does not certify one. Where the answer carries a consequence — a load path, a lifting duty, a pressure boundary, a fastener holding something that can fall — confirm it against the design code that governs the application, and against the manufacturer’s own rating, before relying on it.
Bearing d × D × B (mm)	G = 0.005·D·B (g), D = OUTSIDE diameter	What 0.005·d·B gives (g), D read as the bore	Ratio	Continuous feed rate (g/h)
20 × 47 × 14	3.29	1.40	2.35×	0.0197–0.0329
30 × 62 × 16	4.96	2.40	2.07×	0.0298–0.0496
40 × 80 × 18	7.20	3.60	2.00×	0.0432–0.0720
60 × 110 × 22	12.10	6.60	1.83×	0.0726–0.1210
80 × 140 × 26	18.20	10.40	1.75×	0.1092–0.1820
100 × 180 × 34	30.60	17.00	1.80×	0.1836–0.3060

The viscosity ratio κ = ν/ν₁, and why this page will not compute ν₁

κ is the actual kinematic viscosity at the operating temperature divided by the RATED viscosity the bearing needs at its size and speed. The numerator is a property of your oil and is straightforward. The denominator ν₁ is read from a chart against the mean diameter and the speed, and this page takes it as an input rather than inventing a curve for it — Tribonet’s own guidance on the subject is to “use the bearing manufacturer’s tools, ISO 281 guidance or validated engineering references to estimate the reference viscosity”, which is an admission that there is no formula to quote. κ matters here because it is one of the two arguments of the a_ISO chart on the L10 life page, the other being the contamination-corrected load ratio e_C·C_u/P. So the two pages meet exactly here: this one computes κ, that one uses it. Note that viscosity UNIT conversion — centistokes, SUS, Engler, Redwood — belongs to the converters plugin and is not linked from here; and note that an ISO VG grade number is the viscosity at 40 °C, not at your operating temperature. The same designation can mean different dimensions in different standards families — ANSI against ISO, inch against metric, one national standard against another. The family used here is named beside every figure; check which one your part was made to.
κ	Lubrication condition	What it means
Below 0.1	Boundary lubrication, no film at all	Metal-to-metal contact through the whole rolling track. The a_ISO charts stop here; below κ = 0.1 they do not apply and the life calculation is not meaningful
0.1 to 1	Mixed film — asperities touch	Wear and smearing become possible and an EP or AW additive starts to earn its keep. a_ISO is below 1, so the calculated life is SHORTER than the basic rating life
1 to 4	Full film. The practical target band	The rolling elements are separated from the raceways. a_ISO can exceed 1, so the life is longer than the basic rating life
Above 4	More film than the bearing can use	No further life benefit — the published implementations CAP κ at 4 for exactly this reason — and real costs: churning loss, heat, and at the extreme the rolling elements sliding instead of rolling

The speed factor, the closed form, the temperature rule and the dose

Compute the speed factor before you compute anything else. n·d_m, with d_m = (d+D)/2 and n in rev/min, tells you which kind of problem you have. Below 200,000 the published guidance is an interval of a year or more and the honest answer is often a sealed bearing greased for life with no schedule at all. Between 200,000 and 400,000 you need a plan. Above 400,000 grease is at its limit, the interval is in weeks, and the question becomes an automatic lubricator or oil. This page’s example — a 40 × 80 mm bearing at 1,450 rev/min — gives 87,000, comfortably in the first band.

The interval itself comes from a published closed form, and it matters which way the brackets go. t_f = K[(14×10⁶/(n√d)) − 4d] hours, with K = 10 for a ball bearing, 5 for a cylindrical roller and 1 for a tapered or spherical one. Written that way it goes to ZERO at a finite speed, which is the formula reporting something true: past that point there is no grease interval and you need oil. This batch found a second printing of the same equation with the bracket moved — (14×10⁶ − 4d)/(n√d) — which never goes negative and therefore never tells you that. At the example’s conditions the two differ by twelve per cent; at high speed they differ by everything.

Then halve it for every 15 °C above 70. The chart and the formula are both referred to an outer-ring temperature of 70 °C, and above that the interval falls as 2^(−ΔT/15) — a quarter at 100 °C, an eighth at 115, a sixteenth at 130. The published factors 1.0, 0.5, 0.25 and 0.125 are exactly that expression. It is the most violent term on the page and the one most often left out, and the temperature to use is the OUTER RING’s, which is typically 10 to 20 °C above a housing surface reading. One source disagreement worth knowing: SKF gives a single 15 °C step while at least one published summary of the same material says 10 to 15, and the 10 °C version is twice as harsh at 30 °C above the reference.

The dose is G = 0.005·D·B grams, and D is the OUTSIDE diameter. That is worth labouring: a published description of the same formula says to multiply the BORE by the width, which on a 40 × 80 × 18 mm bearing gives 3.6 g instead of 7.2 — exactly half. The continuous feed rate for a bearing on an automatic lubricator is a different quantity again, (0.3 to 0.5)·D·B×10⁻⁴ grams per hour, and it is not the dose divided by the interval. For the initial fill, the bearing’s own free space can be computed honestly from its catalogue mass — the annulus it occupies, less the steel actually in it — and the HOUSING around it is filled to 30 to 50 per cent of its free space, not to 100.

Over-greasing is one of the commonest causes of bearing failure, and it is always done kindly. Grease that cannot escape is churned by the rolling elements. Churning generates heat. Heat oxidises the base oil and bleeds it out of the thickener. What is left is a hard soap that no longer lubricates, in a bearing that is now running hot. A bearing packed solid runs hotter than one filled correctly — which is why the fill percentage exists, why a relief path for the old grease matters as much as the nipple does, and why a maintenance instruction should say a mass in grams rather than a number of pump strokes (a grease gun’s delivery per stroke varies by a factor of several between guns).

Grease life and the relubrication interval are different quantities. Grease life is how long the grease lasts before it has degraded past usefulness. The relubrication interval is when to add more, and it is conventionally set shorter — roughly half of the grease life in the published treatments — because the point of relubricating is to replace the grease in the track BEFORE it has failed, not after. A sealed bearing has no relubrication interval at all, only a grease life, and once that is spent the bearing is spent with it. Both are separate from the RATING LIFE on the L10 page, which models fatigue and does not see lubricant at all except through a_ISO.

What this page refuses. The rated viscosity ν₁, which is the denominator of κ, comes from a chart read against the mean diameter and the speed, and there is no formula for it — Tribonet’s own guidance on the subject is to use the maker’s tools or ISO 281’s own guidance to estimate it. So ν₁ is an input here and κ is computed from it, exactly as a_ISO is an input on the life page. Viscosity UNIT conversion between centistokes, SUS, Engler and Redwood belongs to the converters plugin and is not linked from here; note in passing that an ISO VG grade number is the viscosity at 40 °C and not at your operating temperature, which is a conversion in its own right.

Frequently asked questions

How often should a bearing be relubricated?

Start from the speed factor n·d_m: below 200,000 the published guidance is a year or more and a sealed greased-for-life bearing is often right; 200,000 to 400,000 is five to twelve months; above 400,000 is weeks and needs an automatic lubricator or oil. Then compute the interval from t_f = K[(14×10⁶/(n√d)) − 4d] hours, and halve it for every 15 °C the outer ring runs above 70. Correct it again for a dirty or vibrating environment. The temperature term is usually the biggest of the three.

How much grease should I put in?

G = 0.005·D·B grams for a relubrication dose, with D the bearing’s OUTSIDE diameter and B its width, both in millimetres. For a 40 × 80 × 18 mm bearing that is 7.2 g. Be careful with the formula: one published description of it says the bore rather than the outside diameter, which gives half the right answer. For an initial fill, the bearing itself is filled completely and the housing around it to 30 to 50 per cent of its free space. State the dose as a MASS on the maintenance instruction, because a grease gun’s delivery per stroke varies by a factor of several between guns.

Why does the interval halve for every 15 °C?

Because grease life is governed by chemistry, not mechanics: the base oil oxidises and the thickener loses its ability to hold it, and both are reaction rates that roughly double for a fixed temperature step. SKF’s published rule is a halving for every 15 °C above 70 °C, and the published correction factors 1.0, 0.5, 0.25 and 0.125 at 70, 85, 100 and 115 °C are exactly 2^(−ΔT/15). Use the OUTER RING temperature, which runs 10 to 20 °C above a housing surface reading.

Can you over-grease a bearing?

Easily, and it is one of the commonest causes of premature failure. Grease that cannot escape gets churned by the rolling elements; churning generates heat; heat oxidises the base oil and bleeds it out of the thickener; what remains is a hard soap in a hot bearing. A bearing packed solid runs hotter than one filled correctly. That is why the housing is filled to 30 to 50 per cent of its free space rather than 100, and why a relief path for the displaced grease matters as much as the nipple does. The failure is usually diagnosed as a lubrication failure, which it is — of the wrong kind.

Is grease life the same as the relubrication interval?

No. Grease life is how long the grease lasts before it has degraded past usefulness; the relubrication interval is when to add more, and it is set shorter — conventionally about half the grease life — because the point is to replace the grease in the rolling track before it has failed rather than after. A sealed bearing has only a grease life and no interval, and when the grease is spent so is the bearing. Neither is the RATING LIFE, which models fatigue and sees lubricant only through the a_ISO factor.

What is the viscosity ratio κ and why is ν₁ an input?

κ = ν/ν₁ is the actual kinematic viscosity at the operating temperature divided by the rated viscosity the bearing needs at its size and speed, and it is the measure of whether there is a lubricant film at all: below 1 asperities touch, 1 to 4 is the target band, and above 4 there is no further benefit (the published implementations cap it at 4). ν₁ is an input because it comes from a chart and there is no honest formula for it — the same reason a_ISO is an input on the life page. κ is one of the two arguments the a_ISO chart is read against, which is where the two pages meet.

The formula gives a negative interval. Is that a bug?

No, it is the formula telling you something. (14×10⁶/(n√d)) − 4d goes negative at a finite speed for any given bore — about 13,800 rev/min for a 40 mm bore — and what that means is that the speed factor has passed the point where a grease relubrication interval exists. The answers are oil, continuous grease feed, or a smaller bearing. It is worth knowing that a second published rendering of the same equation has the bracket in a different place and can never go negative, so it reports a cheerful few hundred hours where the correct form reports none.

Related calculators

References

Machinery Lubrication, “How to change your lubrication culture”, and Reliability Solutions, “How to calculate bearing relubrication intervals”. Between them the source for the closed-form relubrication interval T = K[(14,000,000/(n√d)) − 4d] hours, for the grease dose G = 0.005 D B grams with D the OUTSIDE diameter, and for SKF’s rule that the interval halves for every 15 °C above 70 °C. NOTE: Machinery Lubrication describes the grease dose as “the bore diameter by the bearing width” where every other printing of the same formula uses the OUTSIDE diameter; on a 6208 that is 7.2 g against 3.6, a factor of two.
STLE, Tribology & Lubrication Technology, April 2009, “Lubricant application: grease volumes and frequencies”. The source for the bearing free-space formula (the annulus less the steel the catalogue mass implies) and for the K factors 10, 5 and 1 for ball, cylindrical/needle roller and tapered/spherical roller bearings. NOTE: its rendering of the interval equation moves a bracket, giving (14×10⁶ − 4d) over n√d. That version can never go negative, so it never tells the reader that the speed factor has passed the point where grease alone will do; the other printing’s form is the one used.
Fabrico, “Bearing relubrication intervals” and “Bearing defect frequencies”. The source for the n·d_m speed-factor bands (below 200,000 a year or more, 200,000 to 400,000 five to twelve months, above 400,000 weeks), for the 30 to 50 per cent housing fill, and for the statement that grease life halves for every 10 to 15 °C — a range where SKF gives a single 15, so both are printed.
Tribonet, Rolling bearing lubrication — ISO 281 kappa factor in practice, and Interlub’s κ-value note. The sources for κ = ν/ν₁, for the practical target band of 1 to 4, and — importantly for what this site will not do — for the fact that ν₁ comes from a manufacturer’s chart: “use the bearing manufacturer’s tools, ISO 281 guidance or validated engineering references to estimate the reference viscosity”.
NTN-SNR, Bearing Wizard — Lifetime calculation. A live manufacturer’s calculation tool, and the source for the ISO 281:1990 a₁ column (0.62, 0.53, 0.44, 0.33, 0.21 at 95 to 99 %), which it still publishes. It is also the source for what a_ISO actually depends on — a_ISO = f(e_CC_u/P, κ) — and for its two hard limits: a_ISO ≤ 50, and κ capped at 4 however good the film is. Those two limits are the reason this site takes a_ISO as an input rather than inventing a curve for it.
ISO 281:2007, Rolling bearings — Dynamic load ratings and rating life. Cited by number, not reproduced. What the freely published preview does confirm, in its own introduction, is the thing that matters most here: “the life modification factors for reliability, a₁, have been slightly changed and extended to 99,95 % reliability” — so the a₁ column differs between the 1990 and 2007 editions, and a figure taken from one edition and used with the other is wrong. Both columns are printed on this page and both are attributed. The preview stops before clause 9, so the a₁ table itself was taken from manufacturers’ catalogues, below.
ISO 15243:2017, Rolling bearings — Damage and failures — Terms, characteristics and causes. Cited by number. It is the document that gives the failure modes this batch keeps naming — surface-initiated and subsurface fatigue, smearing, fretting corrosion, false brinelling — their agreed names, which matters because a defect frequency identifies a LOCATION and not a cause.