HEPA efficiency is a constant — H14 means 99.995%, from the day it is installed to the day it is replaced. Chemical filters are entirely different: their efficiency decays over time, and they will inevitably fail. ASHRAE 145.2 answers the question "when will it fail."
Why Particle-Filter Test Methods Do Not Apply to Chemical Filters
Particle filters (HEPA / ULPA / general HVAC) work by physical interception — the media structure is fixed and efficiency is essentially constant (barring extreme clogging or damage). EN 1822 can test efficiency once and it represents the entire service life.
Chemical filters work by chemical adsorption or reaction — impregnated reagents on the activated-carbon surface react with target gases, consuming reaction sites one by one. Like a sponge: it absorbs water quickly at first, slows as it fills, and eventually saturates completely.
This means chemical filter "efficiency" is not a single number but a curve that declines over time. Evaluating a chemical filter requires not "what is the efficiency?" but "under what conditions, how long before efficiency drops to what level?"
ASHRAE 145.2 defines the standardized test method for that "under what conditions."
ASHRAE 145.2 Test Apparatus
ASHRAE 145.2 Test Setup
A chemical filter is mounted in a test duct; a "challenge gas" at known concentration is fed upstream, downstream concentration is measured — removal efficiency over time is calculated
Standard cylinder or permeation tube generates target gas at known concentration (e.g. toluene, SO₂, NH₃)
Challenge gas mixes with clean air in the blending section to ensure uniform upstream concentration
Sample before the chemical filter to confirm actual upstream concentration C_up
The chemical filter under test, installed in the standard test section
Sample after the filter to measure breakthrough concentration C_down and calculate efficiency
Challenge gas concentration is typically 10–100 ppm (far above real-environment ppb levels) to accelerate media saturation within a reasonable test duration. Efficiency = (1 − downstream / upstream) × 100%. "Breakthrough time" is when efficiency first drops below a threshold (usually 50%).
The core concept is simple:
- 1Generate a known concentration of challenge gas — typically toluene (MC representative), SO₂ (MA representative), or NH₃ (MB representative)
- 2Pass it through the chemical filter at a fixed face velocity — simulating actual use conditions
- 3Continuously measure downstream concentration — using PID, GC, or chemical detectors
- 4Plot penetration % vs time
Testing continues until penetration reaches 100% (full saturation) or a predetermined stopping condition (e.g. penetration > 95%).
Key parameters:
| Parameter | Description | Why It Matters |
|---|---|---|
| Challenge gas type | Represents the AMC category you want to filter | Different gases map to different impregnation formulas |
| Challenge concentration | Typically 10–200 ppm (accelerated test) | Real environments are ppb-level; models are needed to extrapolate |
| Face velocity | Airflow speed through the filter | Higher velocity = shorter dwell time = faster breakthrough |
| Temperature / humidity | Test environment conditions | High humidity and heat accelerate breakthrough |
Note: ASHRAE 145.2 test concentrations (ppm range) are far higher than real-use environments (ppb range) to complete the test in a reasonable time. Extrapolating from ppm results to ppb service life requires mathematical tools like the Wheeler-Jonas model.
How to Read a Breakthrough Curve
Typical Breakthrough Curve
Chemical filter efficiency is not constant like HEPA — it degrades over time. The breakthrough curve is the "X-ray" of how efficiency drops
Particle filters (HEPA) maintain near-constant efficiency — you only need to know whether it passes. Chemical filters inevitably decay, so "how long it lasts" is the critical metric. Two filters with 99% initial efficiency might last 6 months vs 2 months — the ASHRAE 145.2 breakthrough curve reveals this difference.
Curve shape depends on media type (activated carbon vs impregnated vs synthetic sorbent), challenge gas, concentration, temperature, and humidity. The same filter will show completely different curves for toluene vs SO₂. ASHRAE 145.2 requires testing to at least the 50% breakthrough point, but many clients require 95%.
The breakthrough curve is the chemical filter's most important performance metric. X-axis: time (or cumulative gas volume). Y-axis: penetration (downstream concentration / upstream concentration × 100%).
The typical shape is an S-curve:
- ▸Early stage: Penetration near 0% — the chemical filter is working well
- ▸Middle stage: Penetration begins climbing — adsorbent is gradually saturating
- ▸Late stage: Penetration approaches 100% — the filter is fully saturated
Key reading points:
| Penetration | Term | Practical Meaning |
|---|---|---|
| 1% | Initial breakthrough | The filter is starting to "leak" but still within tolerance |
| 10% | Early breakthrough | Replacement threshold for some applications (e.g. office HVAC) |
| 50% | Half-life point | The "standard reference point" for filter life; most reports use this |
| 95% | Near saturation | Upper life limit for strict applications (beyond this = total failure) |
50% vs 95% Breakthrough — Which Should the Customer Require?
This is the most common confusion when procuring chemical filters.
50% breakthrough (half-life) is the common reference in ASHRAE 145.2 reports — a report may say "50% breakthrough at 120 minutes," meaning penetration reached 50% at 120 minutes under test conditions.
But 50% does not mean "still usable for 120 minutes" — it means "at this point, half the challenge gas is getting through." For a semiconductor fab, 1% penetration may already be out of spec.
Which breakthrough point to choose depends on your application:
| Application | Suggested Threshold | Rationale |
|---|---|---|
| Semiconductor advanced node | 1% – 5% | ppb tolerance — cannot wait until 50% |
| Semiconductor mature node | 5% – 10% | Wider tolerance |
| Museum / archive storage | 10% – 20% | Long-term protection, not real-time process |
| Office HVAC | 50% | Comfort requirement, not process requirement |
Practical advice: When requesting reports from suppliers, explicitly ask for "time to reach Z% penetration at X ppm challenge and Y m/s face velocity." A single "95% efficiency" number is meaningless — that may be the first-second reading.
Comparison with ISO 10121
ISO 10121 is the international chemical-filter test standard, similar in logic to ASHRAE 145.2 with some differences:
| Item | ASHRAE 145.2 | ISO 10121 |
|---|---|---|
| Publisher | ASHRAE (US) | ISO (International) |
| Challenge gas options | Multiple (toluene, SO₂, NH₃, etc.) | Similar but slightly different list |
| Breakthrough curve | Required | Required |
| Efficiency grading | No grading; reports breakthrough time directly | Has a grading system |
| Pressure-drop test | Included | Included |
ISO 10121 adds an "efficiency grading" concept — converting breakthrough time into grade codes, similar to how ISO 16890 grades particle filters into ePM1 / ePM2.5. The semiconductor industry still primarily uses ASHRAE 145.2, but ISO 10121 reports are acceptable in cross-border procurement.
SEMI F21 → ASHRAE 145.2: The Selection Bridge
SEMI F21 tells you "what AMC categories your fab must control." ASHRAE 145.2 tells you "how long your selected chemical filter will last." The bridging workflow:
- 1Use SEMI F21 to identify AMC categories and tolerances
- ▸Example: MB (bases) < 1 ppb, MC (organics) < 5 ppb
- 1Select the matching chemical filter formula
- ▸MB → acid-impregnated activated carbon
- ▸MC → high-surface-area virgin activated carbon
- 1Use the ASHRAE 145.2 report to confirm service life
- ▸Read the breakthrough curve: at your allowable penetration (e.g. 1%), how many hours does it last?
- ▸Account for differences between test and actual conditions (concentration, velocity, temperature/humidity); use models to convert
- 1Set the replacement cycle
- ▸Breakthrough-curve life × safety factor (typically 0.6–0.8) = actual replacement cycle
- ▸Combine with online monitoring for real-time verification
For chemical filter V-Bank selection architecture, see Chemical Filter V-Bank Selection.
FAQ
Q: Why does a filter with "99% initial efficiency" still break through?
A: 99% is the efficiency at second one after installation. A chemical filter's adsorbent is finite — each molecule adsorbed consumes one site. As sites fill, efficiency continuously declines. Like a new sponge soaking up a puddle — after a few uses it stops absorbing. Breakthrough is inevitable; the question is only "how fast."
Q: Is a report meaningful if it does not name the challenge gas?
A: No. Chemical filters have wildly different adsorption capacities for different gases — the same filter may show 200 minutes to breakthrough for toluene (MC) but only 30 minutes for SO₂ (MA). An "efficiency" number without a specified challenge gas is incomparable. When requesting reports, always specify the actual gas you need to block (or SEMI F21 category).
Q: Why does high humidity accelerate chemical filter breakthrough?
A: Two mechanisms. First, water molecules compete for adsorption sites — activated carbon's capacity is finite, and water occupying sites leaves fewer for the target gas. Second, some impregnants (e.g. KOH) liquefy and wash off in high-humidity environments, directly reducing available reaction sites. ASHRAE 145.2 reports must state test humidity — the same filter at 30% RH vs 80% RH may show a 2–3× life difference.
Q: What is the difference between impregnated and virgin activated carbon?
A: Virgin (non-impregnated) activated carbon relies on physical adsorption — van der Waals forces in micropores capture gas molecules, effective for large organic molecules (MC) but poor for small inorganic gases (MA/MB). Impregnated carbon adds chemical reagents to the surface, relying on chemical reaction — e.g. KOH impregnation reacts with HF/SO₂ to form non-volatile salts, irreversibly "locking" the target gas. The trade-off: impregnation partially blocks micropores, reducing physical adsorption capacity for organics. Hence MA/MB use impregnated carbon; MC uses virgin carbon — each in its role.
Q: My actual environment does not match lab conditions — is the report value still accurate?
A: Not 100%. ASHRAE 145.2 tests use controlled conditions (fixed concentration, velocity, temperature/humidity); real environments fluctuate. Adjustments typically include: (1) use the Wheeler-Jonas model to extrapolate ppm test results to ppb service life; (2) apply temperature/humidity correction factors; (3) multiply by a 0.6–0.8 safety factor. The report value is the upper bound under ideal conditions; actual life will be shorter. This is why advanced fabs do not rely on predictions alone but use online monitoring to decide when to replace in real time.
