On this concept: Formula

Not sure which lesson to do first? Find your level and get a plan →

Math · Advanced Functions · Grade 9-12

Saturation

Saturation is when growth slows down and approaches a limiting value (a ceiling) asymptotically instead of rising forever.

f(x)=L1+e−k(x−x0) (logistic function approaching limit L)

Quick Answer

Saturation is when growth slows down and approaches a limiting value (a ceiling) asymptotically instead of rising forever. You see it whenever growth uses up something that runs out: fish filling a pond with limited food, a rumor reaching a school where most people have already heard it, a new app spreading through a town. Early on the quantity can grow almost exponentially; near the ceiling each period adds less and less. The logistic function f(x)=L1+e−k(x−x0) is the standard model, with L as the ceiling.

Why This Matters

Saturation is what stops an early growth spurt from being projected forever. An app that doubled its users every month for half a year cannot keep doing it once most possible users have joined. A saturation model lets you estimate the ceiling, find when growth is fastest (halfway to L in the logistic model), and avoid forecasts larger than the whole population.

Intuitive Explanation

Imagine a rumor in a school of 1,000 students. At first few people know it, and each tells others, so the number who have heard it roughly doubles each day: 10, 20, 40, 80. But as more students already know, more of each teller's conversations are with people who have heard it. New listeners get scarce, the daily increase shrinks, and the count bends over toward 1,000 without passing it.

That gives the S-shaped logistic curve: a slow start, a steep middle and a flat top. In f(x)=L1+e−k(x−x0), at x=x0 the exponential term is e0=1, so f(x0)=L2: the steepest growth happens exactly halfway to the ceiling. As x grows, e−k(x−x0) shrinks toward 0, the denominator approaches 1, and f(x) approaches L, so y=L is a horizontal asymptote.

Flattening is not the same as falling. A cup of hot coffee also levels off, but it is cooling down toward room temperature: that is decay toward a floor. In saturation the quantity is still increasing at every moment, only by smaller amounts. Look at the direction of each change: increases that keep shrinking mean saturation; decreases that keep shrinking mean decay.

Think of it like this. A theatre filling before a show: while the hall is empty, people stream in through every door; as seats run out, latecomers take longer to find one and the count creeps toward capacity. The picture fits the slowdown near the top, though a real theatre can fill exactly, while a logistic curve only gets closer.

Core idea. A saturating quantity grows quickly while there's room, then slows and flattens as it nears a ceiling it never quite reaches.


When to Use

Use a saturation model when a quantity is growing but held back by a capacity: phrases like levels off, carrying capacity, market saturation or "approaches a maximum of". In data, look for increases that grow at first and then shrink while the values stay below a ceiling. If the increases keep growing with no limit in sight, use exponential growth; if the quantity is falling toward a floor, use exponential decay.

Tip. Ask: Does the quantity grow fast early, then slow and flatten toward a ceiling it never passes?

How to Recognize It

Before choosing a saturation model, look for a limit on how large the quantity can get and check how each period's increase changes as the quantity grows.

  1. Is there a fixed supply the growth uses up, such as the people in a town, the food in a pond or the seats in a hall?

    If yes, the growth must eventually slow, and the size of that supply is a natural first guess for the ceiling L. If nothing runs out, exponential growth may describe it.

  2. Does the question ask when the growth is fastest?

    If yes, for a logistic model the fastest growth comes at the halfway mark, when the quantity equals L2, at x=x0. Before that point each increase is bigger than the last; after it, smaller.

  3. Is the quantity actually getting smaller each period?

    If yes, it is decay, not saturation, even if the drop slows near the end. A saturating quantity keeps rising, only by less each time; its flattening is not a decline.

  4. Does the growth slow down from the very first step, with no fast middle stretch?

    If yes, as with a cold drink warming to room temperature, the model is limited growth, y=L−(L−y0)e−kt, rather than logistic, which starts slowly and speeds up first.

Saturation compared with Exponential growth, Horizontal asymptote, and Exponential decay

All four rows describe curves that bend toward or away from a horizontal line, so they are easy to swap. Ask whether the quantity is rising or falling, and whether a limit is holding it back or nothing is.

Saturation

Meaning
Use this when a growing quantity slows as it approaches a fixed maximum it never exceeds.
Key test
Does the quantity grow fast early, then slow and flatten toward a ceiling it never passes?
Formula
f(x)=L1+e−k(x−x0) (logistic function approaching limit L)
Example
A pond can hold at most L=1000 fish. The population follows a logistic curve, fast early then slowing. What value does it approach over time?

Exponential growth

Meaning
Grows faster and faster with no ceiling.
Key test
Use early on or when no limit constrains the quantity.
Formula
y=a⋅bx, b>1
Example
Unchecked bacteria in unlimited food

Horizontal asymptote

Meaning
The ceiling line y=L the curve approaches but never reaches.
Key test
Use to name the limiting value itself, not the whole growth pattern.
Formula
lim⁡x→∞f(x)=L
Example
y=L as the plateau

Exponential decay

Meaning
Falls toward zero, not up toward a positive ceiling.
Key test
Use when the quantity shrinks each period rather than approaching a max.
Formula
y=a⋅bx, 0<b<1
Example
Cooling coffee, drug clearing the body

Formula & Notation

f(x)=L1+e−k(x−x0) (logistic function approaching limit L)
f(x)=L1+Ce−kx where lim⁡x→∞f(x)=L (carrying capacity) and f′(x)=kf(x) ⁣(1−f(x)L)

How to read it: L denotes the carrying capacity (saturation level). lim⁡x→∞f(x)=L indicates the asymptotic limit.

Worked Examples

Example 1 — Population with a cap

Easy

A pond can hold at most L=1000 fish. The population follows a logistic curve, fast early then slowing. What value does it approach over time?
Show the steps and the answer
  1. Write the model: P(t)=10001+e−k(t−t0) with k>0.

    The pond's limit of 1000 fish is the ceiling L in the logistic formula.

  2. Let t grow large: e−k(t−t0)→0.

    An exponent that keeps getting more negative makes the exponential term shrink toward 0.

  3. So 1+e−k(t−t0)→1 and P(t)→10001=1000.

    The denominator is always slightly more than 1, so the population stays just below 1000.

  4. Check with numbers: if k=1 and t0=0, then P(5)=10001+e−5≈10001.0067≈993.

    Five time units past the midpoint, the population is within 1% of 1000 but still has not reached it.

Answer: Approaches 1000 fish

Takeaway: The ceiling L is the long-run value: the population gets as close to 1000 as you like but never passes it.

Example 2 — No ceiling

Standard

An investment grows 6% per year with no cap. Will it saturate?
Show the steps and the answer
  1. Write the model: A(t)=A0⋅1.06t.

    A fixed 6% gain each year is multiplication by 1.06, pure exponential growth.

  2. Check the yearly gains on 1,000 dollars: 60, then 63.60, then 67.42.

    Each gain is bigger than the last, the opposite of saturation, where the gains shrink.

  3. Look for a ceiling: nothing in the problem caps the balance, and 1.06t eventually passes any number you name.

    Without a limiting factor there is no L for the curve to approach.

Answer: No — it grows without bound

Takeaway: Saturation needs a limit that slows growth; a constant percent gain with no cap keeps speeding up.

Example 3 — Find the mistake: doubling past the ceiling

Application

A new app spreads through a town of 20,000 people. Its users after t months are modeled by U(t)=200001+39⋅2−(t−1), which gives about 500, 976, 1,860 and 3,404 users in months 1 to 4. Ben writes: "The users nearly double every month, so in month 7 there will be about 3404⋅23≈27,000." Find the mistake and correct it.
Show the steps and the answer
  1. Compare Ben's forecast with the town: 27,000 users is more than the 20,000 people who live there.

    The number of users can never exceed the population, so the doubling has to stop.

  2. Name the error: Ben extended the early doubling as if there were no ceiling.

    While users are few, the term 39⋅2−(t−1) is large and the model behaves almost like doubling; that changes as the term shrinks.

  3. Evaluate the model: 2−(7−1)=164, so U(7)=200001+39/64=200001.609375≈12,427.

    The model already contains the slowdown, so substituting gives the honest forecast.

  4. Measure the slowdown: from month 4 to month 7 the users multiply by about 12,427/3,404≈3.7, not 8.

    Growth is still positive but well below doubling: the bend of a saturation curve.

Answer: About 12,427 users in month 7, not 27,000; the count keeps rising toward 20,000 but never passes it.

Takeaway: Early saturation data can look exponential; test any projection against the ceiling before extending a doubling pattern.

Where You'll See It

  • Charging a phone battery

    Fast charging can take a phone from empty to half full in about half an hour, then slows sharply: the charger reduces the current as the battery fills, to protect it. A plot of charge against time climbs steeply and then bends over toward 100%, so the last stretch takes the longest.

  • Fish in a stocked pond

    A pond stocked with 50 fish that can support 1,000 grows quickly at first because food is plentiful. By 900 fish, food is scarce and births barely outpace deaths. Fishery managers aim to keep such a population near the midpoint, 500 here, where the logistic model grows fastest and can replace the most fish caught.

Common Mistakes

  1. The slip: Extrapolating early growth straight up

    Instead: saturation means the rate falls near the limit, so the curve bends over.

  2. The slip: Confusing the ceiling L with the final value being reached exactly

    Instead: the curve approaches L asymptotically, never equaling it.

  3. The slip: Mistaking the flattening for decay

    Instead: in saturation the quantity is still rising, just ever more slowly, not shrinking.


Mini Practice

Try these on your own. Tap Reveal when you want to check.

  1. A population follows f(t)=5001+9e−0.4t. Find the starting value f(0) and the value the population approaches over time.

    Hint: At t=0, e0=1. For large t, what happens to e−0.4t?

  2. For the same model f(t)=5001+9e−0.4t, find f(10) to the nearest whole number.

    Hint: The exponent is −0.4×10=−4.

  3. A rumour spreads by N(t)=20001+199e−t people after t days. How many people know at the start, and when does half the limit know it?

    Hint: Half the limit is 1000, which happens when the denominator is 2.

  4. Users of a new app, in thousands, are 100, 300, 700, 900, 960 and 985 in years 1 to 6. Is this saturating? Estimate the limit.

    Hint: Look at the yearly increases after year 3.

  5. Bacteria in a dish with a fixed amount of food double every hour at first. Over a long time, exponential growth or saturation?

    Hint: What happens when the food starts to run short?

  6. A student sees a logistic curve flatten near its limit L=800 and says 'the population is now decaying.' Find and correct the mistake.

    Hint: Is the population going down, or rising more slowly?

Want the full set?

50 practice questions for this concept — free to try, every one with a complete worked solution showing the why, not just the answer.

Frequently Asked Questions

Why does a logistic curve start slowly if nothing is holding it back yet?

The early growth is slow because the quantity itself is small, not because of the limit. In a logistic model, the gain each period is roughly proportional to the current amount times the room left. With 2 people knowing a rumour, even a juicy one adds only a few listeners a day. As more people spread it, the daily gain rises, until the shrinking room left takes over. Little to start from, then little room left: both make growth slow.

Does a real population ever pass its carrying capacity?

Yes, often for a while. Deer on an island with no predators can overshoot the number the food supports, then crash once the plants are stripped. The smooth logistic curve assumes a population reacts to crowding instantly, but animals breed according to conditions months earlier, so the response lags. That delay produces overshoot and swings above and below the limit. The logistic model describes the average trend, not every rise and fall.

Can the ceiling itself change over time?

Yes, and real models often need to allow for it. The ceiling for a new app in a town is set by the town's population, which grows; a pond's ceiling for fish rises if more food is added and falls in a drought. The basic logistic formula treats LL as fixed, which is fine over a short time. When a quantity levels off and then starts climbing again, the usual reason is that the limit moved.

Why does a phone charge slowly for the last few percent?

Charging a lithium battery follows a saturating curve on purpose. Up to roughly 80%, the charger pushes a steady, high current. Near full, pushing that hard would overheat and wear the cells, so the charger holds the voltage steady and lets the current fall, and the charge creeps toward 100%. The percentage display shows the saturation shape: quick at first, then each extra percent taking longer than the one before.

Learning Path

Saturation

You are here

Next →

Nothing after this yet
Before this you need Asymptote (a line a graph approaches but never reaches) and Growth vs Decay (multiplying by a fixed factor each period). Saturation combines them: growth that starts out looking exponential and bends toward a horizontal asymptote y=L. With it you can model populations, product adoption and learning curves, and later the logistic differential equation describes the same growth slowed by its own size.

See Also