Appendix B

MTBF, FIT, AND PPM CONVERSIONS

One of the most often used numbers in reliability is the Mean Time Between
Failures (MTBF) number. MTBF represents the average time one can expect a
device to operate without failing. There is no assurance that the consumer will
realize this failure free time period because the MTBF is a statistical average. In
fact, if a consumer experiences a failure, the likelihood of an additional failure is
the same before the failure occurs as it is after the failure is repaired (Table B.1).
For example:

If the MTBF = 8,760 h

Then on average a unit will fail every 8,760 h or once a year.
1 year = 356 days × 24 h/day = 8,760

Viewed another way:

If there are 10,000 of these systems in the field, then the manufacturer can
expect 10,000 failures every year (for a repairable system), and if the product
is nonrepairable system, there will be about 6,700 failures. This is covered in
greater detail in the next section.
The Failures In Time (FIT) rate is defined as the failures in time per bil-
lion hours.
It is easy to convert between Mean Time Between Failures (MTBF), Failures
In Time (FIT), and Parts Per Million (PPM) rates.

Mean Time Between Failure (MTBF)

There is a lot of confusion about the term MTBF. When the layperson hears
that a device has an MTBF of 10,000 h, they often think that this means that
this device will not have a failure for at least 10,000 h. This is not the case.
What this means is that for a group or fleet of systems with an MTBF of
10,000 h the average rate of failure will be 10,000 h. Some of the units in this
larger group will actually have a failure rate of the stated MTBF rate, while

Improving Product Reliability: Strategies and Implementation. Mark A. Levin and Ted T. Kalal
 2003 John Wiley & Sons, Ltd ISBN: 0-470-85449-9

299
300 Appendix B

Table B.1 Conversion Tables for FIT to MTBF & PPM

FIT MTBF PPM FIT MTBF PPM
1 1,000,000,000 9 200,000 5,000 1,752,000
2 500,000,000 18 300,000 3,333 2,628,000
3 333,333,333 26 400,000 2,500 3,504,000
4 250,000,000 35 500,000 2,000 4,380,000
5 200,000,000 44 600,000 1,667 5,256,000
6 166,666,667 53 700,000 1,429 6,132,000
7 142,857,143 61 800,000 1,250 7,008,000
8 125,000,000 70 900,000 1,111 7,884,000
9 111,111,111 79 1,000,000 1,000 8,760,000
10 100,000,000 88 1,100,000 909 9,636,000
20 50,000,000 175 1,200,000 833 10,512,000
30 33,333,333 263 1,300,000 769 11,388,000
40 25,000,000 350 1,400,000 714 12,264,000
50 20,000,000 438 1,500,000 667 13,140,000
60 16,666,667 526 1,600,000 625 14,016,000
70 14,285,714 613 1,700,000 588 14,892,000
80 12,500,000 701 1,800,000 556 15,768,000
90 11,111,111 788 1,900,000 526 16,644,000
100 10,000,000 876 2,000,000 500 17,520,000
200 5,000,000 1,752 3,000,000 333 26,280,000
300 3,333,333 2,628 4,000,000 250 35,040,000
400 2,500,000 3,504 5,000,000 200 43,800,000
500 2,000,000 4,380 6,000,000 167 52,560,000
600 1,666,667 5,256 7,000,000 143 61,320,000
700 1,428,571 6,132 8,000,000 125 70,080,000
800 1,250,000 7,008 9,000,000 111 78,840,000
900 1,111,111 7,884 10,000,000 100 87,600,000
1,000 1,000,000 8,760 20,000,000 50.0 175,200,000
2,000 500,000 17,520 30,000,000 33.3 262,800,000
3,000 333,333 26,280 40,000,000 25.0 350,400,000
4,000 250,000 35,040 50,000,000 20.0 438,000,000
5,000 200,000 43,800 60,000,000 16.7 525,600,000
6,000 166,667 52,560 70,000,000 14.3 613,200,000
7,000 142,857 61,320 80,000,000 12.5 700,800,000
8,000 125,000 70,080 90,000,000 11.1 788,400,000
9,000 111,111 78,840 100,000,000 10.0 876,000,000
10,000 100,000 87,600 200,000,000 5.0 1,752,000,000
20,000 50,000 175,200 300,000,000 3.3 2,628,000,000
30,000 33,333 262,800 400,000,000 2.5 3,504,000,000
40,000 25,000 350,400 500,000,000 2.0 4,380,000,000
50,000 20,000 438,000 600,000,000 1.7 5,256,000,000
60,000 16,667 525,600 700,000,000 1.4 6,132,000,000
70,000 14,286 613,200 800,000,000 1.3 7,008,000,000
80,000 12,500 700,800 900,000,000 1.1 7,884,000,000
90,000 11,111 788,400 1,000,000,000 1.0 8,760,000,000
100,000 10,000 876,000
(continued overleaf )
 MTBF, FIT, and PPM conversions 301

Table B.1 (continued)

MTBF FIT PPM MTBF FIT PPM
1 1,000,000,000 8,760,000,000 200,000 5,000 43,800
2 500,000,000 4,380,000,000 300,000 3,333 29,200
3 333,333,333 2,920,000,000 400,000 2,500 21,900
4 250,000,000 2,190,000,000 500,000 2,000 17,520
5 200,000,000 1,752,000,000 600,000 1,667 14,600
6 166,666,667 1,460,000,000 700,000 1,429 12,514
7 142,857,143 1,251,428,571 800,000 1,250 10,950
8 125,000,000 1,095,000,000 900,000 1,111 9,733
9 111,111,111 973,333,333 1,000,000 1,000 8,760
10 100,000,000 876,000,000 1,100,000 909 7,964
20 50,000,000 438,000,000 1,200,000 833 7,300
30 33,333,333 292,000,000 1,300,000 769 6,738
40 25,000,000 219,000,000 1,400,000 714 6,257
50 20,000,000 175,200,000 1,500,000 667 5,840
60 16,666,667 146,000,000 1,600,000 625 5,475
70 14,285,714 125,142,857 1,700,000 588 5,153
80 12,500,000 109,500,000 1,800,000 556 4,867
90 11,111,111 97,333,333 1,900,000 526 4,611
100 10,000,000 87,600,000 2,000,000 500 4,380
200 5,000,000 43,800,000 3,000,000 333 2,920
300 3,333,333 29,200,000 4,000,000 250 2,190
400 2,500,000 21,900,000 5,000,000 200 1,752
500 2,000,000 17,520,000 6,000,000 167 1,460
600 1,666,667 14,600,000 7,000,000 143 1,251
700 1,428,571 12,514,286 8,000,000 125 1,095
800 1,250,000 10,950,000 9,000,000 111 973
900 1,111,111 9,733,333 10,000,000 100 876
1,000 1,000,000 8,760,000 20,000,000 50.0 438
2,000 500,000 4,380,000 30,000,000 33.3 292
3,000 333,333 2,920,000 40,000,000 25.0 219
4,000 250,000 2,190,000 50,000,000 20.0 175
5,000 200,000 1,752,000 60,000,000 16.7 146
6,000 166,667 1,460,000 70,000,000 14.3 125
7,000 142,857 1,251,429 80,000,000 12.5 110
8,000 125,000 1,095,000 90,000,000 11.1 97
9,000 111,111 973,333 100,000,000 10.0 88
10,000 100,000 876,000 200,000,000 5.0 44
20,000 50,000 438,000 300,000,000 3.3 29
30,000 33,333 292,000 400,000,000 2.5 22
40,000 25,000 219,000 500,000,000 2.0 18
50,000 20,000 175,200 600,000,000 1.7 15
60,000 16,667 146,000 700,000,000 1.4 13
70,000 14,286 125,143 800,000,000 1.3 11
80,000 12,500 109,500 900,000,000 1.1 10
90,000 11,111 97,333 1,000,000,000 1.0 9
100,000 10,000 87,600
302 Appendix B

Table B.1 (continued)

PPM MTBF FIT PPM MTBF FIT
1 8,760,000,000 0.1 200,000 43,800 22,831
2 4,380,000,000 0.2 300,000 29,200 34,247
3 2,920,000,000 0.3 400,000 21,900 45,662
4 2,190,000,000 0.5 500,000 17,520 57,078
5 1,752,000,000 0.6 600,000 14,600 68,493
6 1,460,000,000 0.7 700,000 12,514 79,909
7 1,251,428,571 0.8 800,000 10,950 91,324
8 1,095,000,000 0.9 900,000 9,733 102,740
9 973,333,333 1.0 1,000,000 8,760 114,155
10 876,000,000 1.1 1,100,000 7,964 125,571
20 438,000,000 2.3 1,200,000 7,300 136,986
30 292,000,000 3.4 1,300,000 6,738 148,402
40 219,000,000 4.6 1,400,000 6,257 159,817
50 175,200,000 5.7 1,500,000 5,840 171,233
60 146,000,000 6.8 1,600,000 5,475 182,648
70 125,142,857 8.0 1,700,000 5,153 194,064
80 109,500,000 9.1 1,800,000 4,867 205,479
90 97,333,333 10.3 1,900,000 4,611 216,895
100 87,600,000 11.4 2,000,000 4,380 228,311
200 43,800,000 22.8 3,000,000 2,920 342,466
300 29,200,000 34.2 4,000,000 2,190 456,621
400 21,900,000 45.7 5,000,000 1,752 570,776
500 17,520,000 57.1 6,000,000 1,460 684,932
600 14,600,000 68.5 7,000,000 1,251 799,087
700 12,514,286 79.9 8,000,000 1,095 913,242
800 10,950,000 91.3 9,000,000 973 1,027,397
900 9,733,333 103 10,000,000 876 1,141,553
1,000 8,760,000 114 20,000,000 438 2,283,105
2,000 4,380,000 228 30,000,000 292 3,424,658
3,000 2,920,000 342 40,000,000 219 4,566,210
4,000 2,190,000 457 50,000,000 175 5,707,763
5,000 1,752,000 571 60,000,000 146 6,849,315
6,000 1,460,000 685 70,000,000 125 7,990,868
7,000 1,251,429 799 80,000,000 110 9,132,420
8,000 1,095,000 913 90,000,000 97 10,273,973
9,000 973,333 1,027 100,000,000 88 11,415,525
10,000 876,000 1,142 200,000,000 44 22,831,050
20,000 438,000 2,283 300,000,000 29 34,246,575
30,000 292,000 3,425 400,000,000 22 45,662,100
40,000 219,000 4,566 500,000,000 18 57,077,626
50,000 175,200 5,708 600,000,000 15 68,493,151
60,000 146,000 6,849 700,000,000 13 79,908,676
70,000 125,143 7,991 800,000,000 11 91,324,201
80,000 109,500 9,132 900,000,000 10 102,739,726
90,000 97,333 10,274 1,000,000,000 9 114,155,251
100,000 87,600 11,416
 MTBF, FIT, and PPM conversions 303

some will fail sooner, and some later. It is understood that with a population
of units the average or mean failure rate will be the stated MTBF rate.

Reliability Defined: The probability that a product will operate satisfac-

torily for a required amount of time under stated conditions to perform
the function for which it was designed.

Taking an example of 100 units that have a 1,000-h MTBF; let’s find out
more about how many failures there will be, how many units will fail before
the stated 1,000 failure rate, how many after, and how many units will have
more than one failure.
The rate of failure is exponential. Here the expression is:

R(t) = Nε−λt (B.1)

the number still surviving without a failure.

N is the number of units shipped; we will use 100.
ε = 2.718 (or the natural logarithm),
λ is the constant failure rate (in failures per million hours).
t = 1,000 h (for this example).

FIT is sometimes used in place of λ, but it is smaller by three orders

of magnitude, or one failure per billion hours of operation. It is read as
‘‘Failures In a Thousand million’’. Therefore 1,000 λ is one FIT.

λ = 1/MTBF (B.2)

λ and MTBF are inversely related MTBF is Mean Time Between Failures.
So (B.1) becomes:
R(t) = Nε −(t)/(MTBF) (B.3)

Example: Let t = 1,000 h

MTBF = 1,000 h
N = 100 new VCRs or TV sets, or any other type of system
R(t) = 100 × ε−(1000 h)/(1000 h between failures)
= 100 × 2.71−1000/1000
= 100 × 2.71−1
= 100 × 0.37
R(t) = 37 Units ‘‘STILL WORKING WITHOUT A FAILURE’’
304 Appendix B

This also means that 63 units had failures. But in 1,000 h shouldn’t all 100
units have had a failure? No; but there still were 100 failures!
At first it seems impossible that there were 100 failures and 37 units were still
working; but the answer is that of the 63 units that had failures, some had more
than one failure. Some had two or there or even more failures. That’s where
the total of 100 failures comes from. The only way this could happen is when
one unit fails, it is quickly repaired and placed back into service. Even after one
failure, as soon as it is repaired there are 100 units that are operating that all
have an MTBF of 1,000 h. Even after 25 or 50 or 63 failures, as soon as that last
failure was repaired there were always 100 units operating; all with an MTBF
of 1,000 h. This is considered the number of failures in ‘‘repairable’’ systems.

Estimating Field Failures

Suppose a product has a 1,000-h system MTBF. Then λ will be 1/1,000 or
0.001. This means that every 1,000 h a system will have a failure. With 100
systems then there will be 100 failures in those 1,000 h. Remember that these
failures will show up in only 63 units; the other 37 units will exhibit no failures
during this time period.
So how many of the 63 units had 1, 2, 3, . . .. or n failures?
The number of units having more than one failure can be determined using:

P (n) = [(λn × t n )/n!] × ε−λt (B.4)

Where:
P (n) is the percent of units exhibiting n failures.
t is the time duration,
n is the number of failures in a single system, (e.g. 1, 2, 3, . . . n).
Let’s learn how many units will have 1, then 2, then 3, and so on, failures
per unit in the group of 63 units that will exhibit these 100 failures.
But first a short refresher in factorials:
Note: 0! Is defined as equaling 1; and 0! is read as ‘‘zero factorial.’’ See
Table B2 for a list of common factorials.
For zero failures: (this is the group of 37 units that had no failures in 1,000 h.)

P (0) = [(0.0010 × 1,0000 )/1] × 2.71−(0.001×1000)

= [(1 × 1)]/1 × 2.71−1
= 1 × 0.37
P (0) = 0.37 or 37%
 MTBF, FIT, and PPM conversions 305

Table B.2 Factorials

n n! n factorial The math

0 0! Zero factorial Defined as 0

1 1! One factorial 1×1=1
2 2! Two factorial 1×2=2
3 3! Three factorial 1×2×3=6
4 4! Four factorial 1×2 × 3 × 4 = 24
5 5! Five factorial 1×2 × 3 × 4 × 5 = 120

So with 100 units there will be 37 units exhibiting zero failures in one MTBF
time period.
How many units will have one failure in 1,000 h?
Substitute 1 for n

P (1) = [(0.0011 × 1,0001 )/1] × 2.71−1

P (1) = (1/1) × 0.37
P (1) = 0.37, or 37% will exhibit one failure.

So with 100 units there will be 37 units exhibiting one failure in one MTBF
time period (1,000 h).
How many units will have two failures in 1,000 h?

P (2) = [(0.0012 × 1,0002 )]/2 × 0.37

P (2) = (1/2) × 0.37
P (2) = 18%

So with 100 units there will be 18 units exhibiting two failures in one MTBF
time period (1,000 h).

P (3) = 6 units exhibiting 3 failures in one MTBF.

P (4) = 1 units exhibiting 4 failures in one MTBF.
P (5) = may be 1 unit exhibiting 5 failures in one MTBF
(numbers are rounded).

A more simple way of finding the percentage of failures encountered in a given

time period is:
P (f ) = λt (B.5)
306 Appendix B

Table B.3 Repairable versus Nonrepairable Systems Still Operating (in MTBF Time Units)

MTBF time periods

1 MTBF 2 MTBFs 3 MTBFs

# Still # Still # Still

# Fails operating # Fails operating # Fails operating

Repairable 100 100 100 100 100 100

systems
Nonrepairable 63 37 86 14 95 5
systems

Find how many will fail in one hundredth of an MTBF time period.

P (f ) = 0.001 × 1,000/100 h
P (f ) = 0.001 × 10
P (f ) = 0.01 or 1%

Using 100 units this means that 1 unit exhibits the very first failure in 10 h. So
the time to first failure is 10 h!!!!!
Which one it will be in the 100 units is a mystery, however . . ..
Interestingly enough one unit will last for 5,000 h before it finally has its
first failure.
Note: These failures have been considered where the failure rate was expo-
nential. There are other failure rates that are Weibull, Log Normal, and more.

Comparing Repairable to Nonrepairable Systems

If the system is comprised of nonrepairable systems, the number of failures in
one MTBF period is lower.
In repairable systems, when a unit fails it is quickly repaired and placed back
into service. If the unit cannot be repaired, then when one unit fails in 100
systems there will be 99 units operating after the first failure. Then 98, 97,
and so on until they all eventually fail. In 100 systems that are nonrepairable,
there will be 67 units that will fail in one MTBF time period. If the MTBF
were 10,000 h, then 2 MTBFs would be 20,000 h. The comparison between
repairable systems and nonrepairable systems is shown in Table B.3.