1.11 · AHL · HL onlyPaper 1Paper 2Paper 3

Infinite geometric series

In a minute

Some go on for ever, yet their running totals settle on one number, the . It exists when ∣r∣<1\lvert r\rvert < 1, and then S∞=u11−rS_{\infty} = \frac{u_{1}}{1 - r}. You use it for long-run totals, for the a model never passes, and for and repeated games.

You'll be able to

  • Explain why partial sums settle on a only when ∣r∣<1\lvert r\rvert < 1.
  • Find the , or work back from it to u1u_{1} or rr.
  • Use it for long-run totals and limits in a model, find when a total passes a value, and comment on the model.
  • Use it for the area of a and for repeated rounds in a Markov chain.
Before you start2 quick questions on what this builds on
Check yourself
Find the sum of the first 44 terms of the geometric series 64+16+4+…64 + 16 + 4 + \dots
Check yourself
Work out 0.7n0.7^{n} for n=5n = 5, 1010 and 2020 on your GDC. As nn gets larger and larger, which value does 0.7n0.7^{n} get closer to?

Revise first: 1.2 Arithmetic sequences and series, 1.3 Geometric sequences and series and 5.1 Limits and the derivative as a rate of change

Adding for ever: the sum to infinity

Eat half a cake, then half of what is left, and so on. Your running totals are 0.50.5, 0.750.75, 0.8750.875, … of the cake. You never pass 11 cake, so the endless sum 12+14+18+…\frac{1}{2} + \frac{1}{4} + \frac{1}{8} + \dots is finite.

These totals S1,S2,S3,…S_{1}, S_{2}, S_{3}, \dots are called partial sums. As in 5.1, they have a : they get closer and closer to 11. We write lim⁡n→∞Sn=1\lim_{n \to \infty} S_{n} = 1.

Why do they settle? In 1.3 you met Sn=u1(1−rn)1−rS_{n} = \frac{u_{1}(1 - r^{n})}{1 - r}. When −1<r<1-1 < r < 1, rnr^{n} gets closer and closer to 00 as nn grows. So SnS_{n} gets closer and closer to u11−r\frac{u_{1}}{1 - r}.

In the formula booklet
S∞=u11−r,∣r∣<1S_{\infty} = \frac{u_{1}}{1 - r}, \quad \lvert r\rvert < 1

It's in the formula booklet. You must check the condition: ∣r∣<1\lvert r\rvert < 1 means −1<r<1-1 < r < 1. Otherwise the terms never shrink in size, so the totals never settle and there is no sum to infinity.

For 5+6+7.2+…5 + 6 + 7.2 + \dots (r=1.2r = 1.2) the totals grow without limit. The formula's 51−1.2=−25\frac{5}{1 - 1.2} = -25 means nothing.

Worked exampleA pendulum's bob travels 4848 cm on its first swing, and each swing is 85%85\% as long as the one before. Find the total distance the model predicts, and interpret it.

Try it first, then reveal one step at a time.

  1. The swing lengths form a with u1=48u_{1} = 48 and r=0.85r = 0.85. As ∣r∣<1\lvert r\rvert < 1, the sum to infinity exists.
0 5 10 15 20 0 50 100 150 200 250 300 350 n (number of swings) Total distance Sn (cm) S∞ = 320 48.0 178.0 257.0
The partial sums of 48+40.8+34.68+…48 + 40.8 + 34.68 + \dots approach S∞=320S_{\infty} = 320 but never reach it.
Check yourself
Find the sum to infinity of 72+54+40.5+…72 + 54 + 40.5 + \dots
Check yourself
Does the geometric series 6−9+13.5−…6 - 9 + 13.5 - \dots have a sum to infinity?

Know two of S∞S_{\infty}, u1u_{1} and rr, and the formula gives the third. Any rr you find needs ∣r∣<1\lvert r\rvert < 1 and must fit the context.

Worked exampleA charity models weekly donations to an appeal as a geometric sequence. The first two weeks raise 3200032000 dollars between them. The model predicts that the appeal will raise 5000050000 dollars in the long run. Find the amount raised in the first week.

Try it first, then reveal one step at a time.

  1. Write both facts as equations: u1+u1r=32000u_{1} + u_{1}r = 32000 and u11−r=50000\frac{u_{1}}{1 - r} = 50000.

Modelling with a sum to infinity

In a model, S∞S_{\infty} is the long-run total. When every term is positive, the total approaches it but never passes it. To find when a total passes a value, use SnS_{n} on your GDC.

Worked exampleA puppy weighs 3.23.2 kg. Next month it gains 1.51.5 kg, and a vet models each later month's gain as 75%75\% of the month before. Find the weight the model says the puppy will never exceed, and comment on the model.

Try it first, then reveal one step at a time.

  1. The monthly gains form a geometric sequence: u1=1.5u_{1} = 1.5 and r=0.75r = 0.75, with ∣r∣<1\lvert r\rvert < 1.
Worked exampleA new app is downloaded 24002400 times in its first week. Each later week, new downloads are 90%90\% of the week before. Use your GDC to find the week in which total downloads first pass 2000020000.

Try it first, then reveal one step at a time.

  1. The weekly downloads are geometric with u1=2400u_{1} = 2400 and r=0.9r = 0.9. Check 2000020000 can be reached: S∞=24001−0.9=24000>20000S_{\infty} = \frac{2400}{1 - 0.9} = 24000 > 20000.
Check yourself
An oil well produces 5000050000 barrels in its first year. Each year it produces 88%88\% of the year before. Find the total the model predicts the well will ever produce. Give your answer to 3 significant figures.
Check yourself
Which is the best comment on the well model's total?

Common mistakes

  • Putting a large nn, or ∞\infty, into SnS_{n} for the most a total can be. Use S∞=u11−rS_{\infty} = \frac{u_{1}}{1 - r} and show the substitution.
  • Forgetting the starting amount: the puppy's limit is 3.23.2 kg plus S∞S_{\infty} of the gains.
  • Using the formula when ∣r∣≥1\lvert r\rvert \ge 1. A such as 5+6+7.2+…5 + 6 + 7.2 + \dots has no sum to infinity.
  • Rounding the crossing point n=17.005…n = 17.005\ldots down. Test S17S_{17} and S18S_{18}, and give 1818.
  • A limitation that ignores the context. Say what breaks the fixed ratio, such as the process stopping.

Exam tips

  • Find the value a total of positive terms with 0<r<10 < r < 1 never exceeds: write u11−r\frac{u_{1}}{1 - r} with the numbers in, add any starting amount, and give units.
  • Determine whether a total of positive terms with 0<r<10 < r < 1 reaches a target: it does only if the target is less than S∞S_{\infty}. Compare them, then answer in words, in context.
  • Justify a rejected ratio: say ∣r∣≥1\lvert r\rvert \ge 1 gives no sum to infinity, or say why the context rules it out.

Key takeaways

In bookletS∞=u11−r,∣r∣<1S_{\infty} = \frac{u_{1}}{1 - r}, \quad \lvert r\rvert < 1
Learn thisA geometric series has a sum to infinity only when −1<r<1-1 < r < 1. Then S∞=lim⁡n→∞SnS_{\infty} = \lim_{n \to \infty} S_{n}.
Learn thisWorking back from S∞S_{\infty} can give two ratios. Reject any with ∣r∣≥1\lvert r\rvert \ge 1, and any the context rules out, and say why.
Learn thisA starting amount plus increases with ratio 0<r<10 < r < 1 never exceeds the starting amount plus S∞S_{\infty} of the increases.
Learn thisLeast nn for a total of positive terms to pass a target: if 0<r<10 < r < 1, first check the target is below S∞S_{\infty}, or it is never passed. Then solve Sn>targetS_{n} > \text{target} on the GDC, test the whole numbers either side and give the first that works.
Learn thisFractals: positive amounts added stage by stage with ratio ∣r∣<1\lvert r\rvert < 1 have a finite total; with r≥1r \ge 1, like the Koch perimeter's 43\frac{4}{3}, they grow without limit.

Practise this

10 past-paper questions on Sum to infinity in the AI HL bank, each with a mark scheme and a worked explanation.

Practise Sum to infinity →