PSLE Science Reality Lab Vol No.175 | “HEPA Filter: 99.97%” — Will the Whole Room Become 99.97% Particle-Free?

PSLE-SCI-REALITY-0175

Wait, What? The Box Says “HEPA 99.97%” — Will 99.97% of the Particles in My Room Disappear?

A fictional air-cleaner package shows a large badge:

HEPA FILTER — 99.97%

A learner reads the number as a room result: “If I turn this machine on, 99.97% of the particles in the whole room must disappear.”

The number can be meaningful without meaning that.

The U.S. Environmental Protection Agency describes a high-efficiency particulate air, or HEPA, filter as a pleated mechanical air filter capable, in theory, of removing at least 99.97% of dust, pollen, mould, bacteria and airborne particles with a size of 0.3 micrometres. That statement describes filter performance under specified conditions. A room is a larger system. Its outcome also depends on how much room air actually passes through the filter, whether air leaks around it, the cleaner’s airflow, where the cleaner sits, how long it runs, room size and mixing, and whether particles continue entering or being generated.

Reality Lab habit: a component efficiency is not automatically the efficiency of the whole system that contains the component.

Quick Answer

The 99.97% HEPA figure is a filter-performance statement, not a promise that an entire room becomes 99.97% particle-free.
The filter can only capture particles carried through it; particles that never enter the cleaner are not captured by that pass.
Airflow matters. A highly efficient filter moving very little air can clean a room more slowly than a well-designed system moving more clean air through the space.
Bypass or leakage around a poorly fitted filter can reduce whole-device performance.
New particles can enter through doors, windows, outdoor air or indoor activities while cleaning continues.
Room size, placement, run time and air mixing affect how quickly particle concentration changes.
A filter rating does not by itself provide personalised health advice or guarantee removal of gases and every pollutant type.

The Exact Learner Job This Volume Owns

This volume owns one narrow real-world evidence-transfer job: how to evaluate a HEPA-filter or air-cleaner claim without turning a filter test efficiency into a whole-room removal percentage.

It does not become the canonical lesson on filtration, respiratory health, fluid flow, particle physics, product certification or environmental health. Those concepts remain with their appropriate scientific owners and public-health authorities. Reality Lab focuses on one communication object: a large percentage badge whose scope can easily be stretched beyond what was tested.

Rebuild the Evidence Object: Test Bench Versus Real Room

Picture two very different scenes.

Scene A: the filter test

Air containing test particles is directed through a filter under a defined procedure. Instruments compare particle amounts before and after the filter. The result describes how efficiently the filter captures particles in that test.

Scene B: the bedroom

The cleaner stands near a wall. The room has furniture, corners, a door that opens, perhaps a window gap, people moving, fabrics releasing particles and air that does not mix perfectly. Only some of the room air enters the cleaner during each minute.

The same filter can perform well in Scene A while the rate of change in Scene B depends on the entire air-cleaning system. The room outcome is therefore not obtained by simply writing “99.97%” beside the room.

The Evidence Chain: Source → Room Air → Cleaner → Filter → Room Air Again

To evaluate a whole-room claim, follow the path of the particles:

Particles enter or are produced in the room.
Air currents move some particles through the space.
A fraction of the room air reaches the cleaner intake.
The fan moves that air through the device.
The filter captures a fraction of particles in the air that passes through it.
Cleaned air returns to the room.
The cleaner repeats the cycle while new particles may continue entering or being generated.

The filter-efficiency number belongs mainly to Step 5. A statement about the concentration in the whole room depends on the entire chain.

Observed, Tested, Claimed and Inferred

Tested: the filter removes at least the stated fraction of specified test particles under the relevant test definition.
Supported device claim: the device uses a HEPA filter meeting that filter definition, if the filter and installation genuinely meet the stated specification.
Possible room inference: running a suitably sized cleaner can reduce airborne particle concentration over time.
Unsupported leap: 99.97% of all particles in the entire room disappear after one pass or one minute.
Unsupported leap: the room becomes completely particle-free.
Unsupported leap: every gas, odour or chemical pollutant is removed by a particle filter.
Unsupported leap: one filter percentage tells a particular person what health outcome they will experience.

The Boundary Check: Filter, Device or Room?

A percentage can belong to different system boundaries. Before accepting a claim, label the boundary:

Boundary	Question	Typical evidence needed
Filter material	What fraction of test particles passing through the filter is captured?	Standardised filter test
Whole cleaner	How much particle-free or particle-reduced air does the device deliver?	Device airflow and performance test
Room	How quickly does airborne particle concentration fall in this space?	Room size, mixing, sources, cleaner placement, run time and measured concentration

The larger the boundary, the more pathways the evidence must cover.

Airflow Check: A Perfect Catcher That Sees No Air Cleans Nothing

Imagine an imaginary filter that captures 100% of particles passing through it but is sealed inside a box with no fan and no airflow. Its filter efficiency is excellent. Its room-cleaning effect is zero because no room air reaches the filter.

This extreme example shows why airflow belongs beside efficiency. EPA guidance on portable air cleaners uses Clean Air Delivery Rate, or CADR, as a useful measure that incorporates both removal efficiency and airflow. A device intended for a larger room generally needs a larger CADR to clean the air at a useful rate.

Original Worked Case 1: Same HEPA Filter, Different Fan Speeds

A fictional cleaner uses the same HEPA filter at two fan settings.

Setting	Filter type	Airflow through cleaner
Low	HEPA	80 m³/h
High	HEPA	240 m³/h

A learner says both settings must clean the room at exactly the same speed because the HEPA filter percentage is unchanged.

Repair: filter efficiency may be similar, but high mode processes more room air per hour. Whole-room particle reduction can therefore occur more quickly, provided other conditions are comparable.

Original Worked Case 2: The Bypass Gap

Cleaner P seals its filter tightly. Cleaner Q has a gap around the filter frame, allowing some air to pass around rather than through the filter.

Both contain the same filter material. Yet Q can deliver worse whole-device particle removal because some air bypasses the filter.

Evidence lesson: the component specification cannot describe paths that avoid the component.

Original Worked Case 3: Small Room Versus Large Hall

The same cleaner runs for one hour in a 12 m² bedroom and a 100 m² hall of similar ceiling height. A post claims, “It removes 99.97% in one hour in any room.”

Repair: room volume changes how many times the cleaner can process an amount of air comparable to the room volume. A device sized for a bedroom may be too small to produce the same concentration change in a much larger hall during the same time.

Original Worked Case 4: Cleaner On, Window Open

A cleaner runs beside an open window during a hazy outdoor period. Indoor particle concentration falls only slightly. The learner concludes that the HEPA filter “failed”.

Repair: outdoor particles may be continuously entering while the cleaner removes particles. The observed concentration is the result of competing processes. Test the filter or device under controlled conditions before blaming the filter material.

Original Worked Case 5: One Sensor Beside the Cleaner

A particle sensor placed directly in the clean-air outlet shows a large drop. A headline says, “The whole room improved by the same amount.”

Repair: the sensor location samples air near the outlet, not necessarily the whole room. Representative room claims need measurements at suitable positions and times.

Original Worked Case 6: The Birthday-Candle Test

In a fictional demonstration, candles are blown out in a room while the cleaner runs. Particle concentration first spikes and later falls. A video cuts away during the first five minutes and shows only the later decline.

Repair: evaluate the complete time series. The cleaner can reduce particles while the edited video still exaggerates how quickly or smoothly the change occurred.

Original Worked Case 7: “99.97% of All Pollutants”

A social-media caption changes “airborne particles” into “all pollutants”.

Repair: HEPA filters are particle filters. Gases and many vapours require different removal mechanisms such as suitable sorbent media. A claim that broadens particles into every pollutant changes the scientific object.

Why 0.3 µm Does Not Mean “Only 0.3 µm Particles”

EPA’s HEPA definition highlights 0.3-micrometre particles. A learner may assume the filter works only for particles exactly that size. That is incorrect. EPA notes that particles larger or smaller than 0.3 µm can be trapped with even higher efficiency in many HEPA filtration conditions.

The important Reality Lab point is not to memorise a filtration curve. It is to avoid converting a test reference size into the only size the filter can capture.

Comparison Check: Two Cleaners Both Say “HEPA”

Cleaner A and Cleaner B both advertise HEPA filters. Does that prove equal room performance?

No. Compare the whole device and use case:

verified filter specification;
airflow or CADR;
recommended room size;
fan speed used during the comparison;
filter fit and bypass control;
noise level if it affects whether users actually run the device at the tested setting;
filter loading and replacement condition;
room placement and air mixing.

Two products can contain filters with similar capture efficiency and still deliver different amounts of cleaned air to a room.

Baseline Check: “Removes 90%” Compared With Which Starting Point?

A product video says indoor particles “fell by 90%”. Ask:

90% relative to the concentration at what starting time?
How long did the device run?
Was the room sealed or ventilated?
Were particle sources stopped?
Where was the sensor placed?
Was there a no-cleaner comparison?
Was the decline greater than natural settling or ventilation alone?

A large before-and-after reduction can be genuine while still needing a comparison to identify how much of the change was caused by the cleaner.

Alternative Explanations for a Falling Particle Reading

If a sensor reading falls after a cleaner is switched on, the cleaner is a plausible cause. But a careful investigation considers other changes that happened at the same time:

the particle source stopped;
particles settled onto surfaces;
a door or window was opened or closed;
outdoor conditions changed;
the sensor was moved;
humidity affected an optical particle sensor;
people left the room and stopped resuspending dust.

A fairer test holds these factors steady or measures them so the cleaner’s contribution can be separated.

What Evidence Would Strengthen a Whole-Room Cleaning Claim?

A verified HEPA filter or clearly specified particle-removal method.
Whole-device airflow or CADR measured under known conditions.
A room size matched to the device’s tested performance.
Particle measurements at representative locations.
A stated run time and fan speed.
A comparison condition without the cleaner or with the cleaner off.
Control of doors, windows and particle sources.
Repeated trials showing a consistent effect.
A full time series rather than one selected before-and-after pair.

What Would Weaken the Claim?

The filter’s 99.97% figure is presented as the room-removal percentage.
No airflow or room size is given.
Air can visibly bypass the filter.
The sensor sits only in the clean-air outlet.
The room has an uncontrolled continuing particle source.
The comparison uses different fan speeds or different room conditions.
The demonstration omits the early part of the time series.
A particle filter claim is expanded to gases or every pollutant.

Tempting Reasoning That Fails

Filter efficiency = room efficiency. Different system boundaries.
99.97% = particle-free room. New particles can enter and some air has not yet passed through the filter.
Same HEPA label = same cleaner performance. Airflow and device design matter.
One outlet sensor = whole room. Sampling position matters.
Low particle reading = every pollutant removed. Particle concentration does not measure gases and every chemical.
One demonstration = universal result. Room size, sources, placement and run time affect transfer.

Model and Measurement Limits

An air cleaner simplifies a complex room into a repeated process: take in air, remove some particles, return air. Models often describe this with airflow, clean-air delivery and room volume. Those models are useful, but real rooms contain imperfect mixing, surfaces that capture and release particles, open doors, people and changing sources.

A particle sensor also measures only what its method can detect in the sampled air. A low-cost optical sensor may respond differently to particle size, composition and humidity than a reference monitor. The filter claim and the sensor claim both need their own measurement boundaries.

How Far Can the Conclusion Travel?

Suppose a device genuinely contains a HEPA filter meeting the EPA definition. A bounded conclusion is:

The filter is designed to capture at least the stated proportion of specified test particles from air that passes through the filter under the relevant test conditions.

The same evidence alone does not justify:

99.97% of particles in the entire room disappear immediately;
the room becomes particle-free;
every pollutant is removed;
the device performs equally in every room size;
there is no bypass or leakage;
a particular person will obtain a specific health outcome.

PSLE-Style Transfer Case: The Two Boxes

Two identical sealed boxes contain the same starting particle concentration. Each receives the same filter material, but Box P has a fan moving 30 litres of air per minute through the filter while Box Q has a fan moving 90 litres per minute. No new particles are added.

A learner writes, “The boxes must clean at the same rate because the filter efficiency is the same.”

Explained answer: the filter material may capture the same fraction of particles from each portion of air that passes through it, but Q processes more air per minute. If other conditions are equal, Q can reduce the box’s particle concentration faster.

Changed-Problem Transfer: A Sieve in a Water Tank

Imagine a sieve that catches 99% of beads in water passing through it. If only one cup of water from a large tank passes through the sieve, has 99% of the beads in the whole tank been removed? No. The sieve efficiency belongs to the processed flow; the tank outcome depends on how much tank water passes through and how often.

The analogy is not an air-cleaner design model, but it preserves the evidence logic: component capture × processed flow × time determines far more than component capture alone.

Delayed Independent Return: Component, Flow, System

Component: what can this part do to material that reaches it?
Flow: how much material actually passes through the part?
System: what other sources, leaks, pathways and storage exist?

Try the same questions later on a water filter, radiator, solar panel, catalytic converter or school ventilation system. A component specification becomes useful only when it is placed back inside the system.

Explained Practice

1. Does 99.97% HEPA efficiency mean 99.97% of the room particles disappear instantly? No. The percentage refers to filter capture under defined conditions for air passing through the filter.

2. Why does fan airflow matter? A filter cannot remove particles from air that never passes through it. More processed air can increase the rate of whole-room cleaning when other conditions are suitable.

3. Why can two HEPA cleaners perform differently? Their airflow, filter fit, bypass leakage, room-size suitability and device design can differ.

4. Can a HEPA particle filter remove every gas? No. Particle capture and gas removal are different scientific jobs.

5. What is the best evidence for a whole-room claim? Measurements of the room concentration over time under controlled, representative conditions, together with verified device performance and room information.

Parent and Tutor Teaching Guide: Draw the Boundary Before Reading the Percentage

Draw three nested rectangles labelled filter, air-cleaner device and room. Put “99.97%” inside only the filter rectangle. Ask the learner what extra information is needed before moving that number into the device rectangle, then the room rectangle.

Useful answers include airflow, bypass, room volume, run time, placement and continuing particle sources. The exact engineering calculation is not the teaching target. The target is recognising that a claim changes when the system boundary changes.

For a second round, use a fictional water filter labelled “removes 95% per pass”. Ask whether one filtered cup proves that 95% of material in an entire swimming pool has disappeared. The child should now transfer the same boundary logic without relying on the word HEPA.

Why This Belongs in PSLE Science Reasoning

The 2026 PSLE Science assessment objectives include interpreting and analysing information, evaluating observations, information and methods, and communicating explanations and reasoning. The 2023 Primary Science syllabus also develops healthy scepticism, evidence-based evaluation, fair comparison and the ability to understand science communicated in different representations.

A HEPA badge is excellent transfer practice because the printed percentage can be scientifically valid while a broader advertising interpretation is not. The disciplined learner asks which object was tested, what air passed through it, what boundary the percentage belongs to and what extra evidence is needed for the whole-room claim.

Authoritative Sources

The Quiet Return

The percentage belongs to the filter test before it belongs to anything larger.

Find the boundary. Follow the air. Then decide how far the claim can travel.

Small Group Tutorials