ULPA vs. HEPA filter: the short answer
A HEPA filter under EN 1822 typically refers to H13 or H14. A ULPA filter refers to U15, U16, or U17. The jump from H14 to U15 looks small on paper, but in penetration terms it is a major step. H14 allows a maximum overall penetration of 0.005%, while U15 reduces that to 0.0005%. That is a 10× reduction in allowable penetration.
That is the first thing buyers should remember:
•H14 = HEPA
•U15-U17 = ULPA
•The test basis is MPPS, not just a simple 0.3 μm marketing statement
Higher efficiency usually means tighter contamination control, but also more system sensitivity to airflow design and Pressure Drop
U15, U16, U17 explained
EN 1822 efficiency classes
Under EN 1822, the commonly referenced high-efficiency classes are:
•H13: ≥ 99.95% overall efficiency
•H14: ≥ 99.995% overall efficiency
•U15: ≥ 99.9995% overall efficiency
•U16: ≥ 99.99995% overall efficiency
•U17: ≥ 99.999995% overall efficiency
The same class table also includes local efficiency / penetration limits, which matter when leak scanning is part of the acceptance requirement.
Why buyers often see "99.9995% @ 0.12 μm"
This is where confusion starts.
In the market, many ULPA datasheets still describe performance as 99.9995% at 0.12 μm. AAF product literature does this directly for ULPA modules and box filters. That wording is still common and familiar to many engineers. But the formal classification logic under EN 1822 / ISO 29463 is based on MPPS testing. That means the safest way to compare products is:
•Check the class first: H14, U15, U16, U17
•Confirm the test standard: EN 1822 / ISO 29463
•Then review how the supplier states the measured efficiency in the datasheet
What the U15 jump really means
Our engineers often see buyers treat U15 as "just a little better than H14." It is better than that. The allowable total penetration drops by one decimal place. In cleanroom terms, that matters when the process is already fighting at the defect level, not just at the room-comfort level.
Why operating rooms usually stop at HEPA, while semiconductor fabs often move to ULPA
Healthcare and operating rooms: the target is infection control, not wafer yield
In healthcare design, the main concern is controlling airborne infectious risk, surgical-site contamination, pressure relationships, and clean airflow over the critical zone. ASHRAE's healthcare guidance states that HEPA filters are required by Standard 170 only for protective-environment rooms, not across every healthcare space. For operating rooms, the emphasis is on the supply diffuser array, airflow pattern, and room pressurization. That is why healthcare discussions usually center on HEPA, not ULPA.
There are healthcare products that offer optional ULPA performance. Camfil's operating-theatre ventilation literature mentions optional ULPA filters for some systems. But that does not change the broader rule: the standard healthcare benchmark is HEPA-grade filtration, while ULPA is a special case, not the default answer.
Semiconductor fabs: the target is defect control at much smaller scales
Semiconductor cleanrooms are different. Camfil's semiconductor guidance describes fabs targeting ISO Class 1 particle levels along with strict control of airborne molecular contamination (AMC). It also highlights the use of HEPA and ULPA filters in applications such as lithography, etch, diffusion, metallization, ion implantation, inspection tools, and reticle or wafer storage. MANN+HUMMEL also describes FFU systems for cleanrooms using H14 to U17 elements.
That is why semiconductor buyers often move from HEPA to ULPA:
•critical particle sizes are smaller
•yield loss is tied directly to nanoscale contamination
•tools and mini-environments often need tighter control than the surrounding fab
•airborne molecular contamination matters alongside particles
Simple rule:
•If the room is protecting a patient and staff workflow, HEPA is usually the reference point.
•If the room is protecting a wafer, mask, reticle, or advanced process tool, ULPA becomes much more common
Does ULPA always beat HEPA?
No. Not automatically.
A better filter on paper is not always the better engineering choice in the field.
Buyers should also compare:
•Initial Resistance
•allowable final Pressure Drop
•fan capacity and energy cost
•airflow uniformity
•sealing and scan-test requirements
•outgassing behavior
•application risk if the filter clogs faster or forces a redesign of the air system
Our engineers often see ULPA specified where H14 would have met the real process requirement with lower system resistance and lower operating cost. The opposite mistake also happens: HEPA gets copied into a semiconductor or microelectronics RFQ simply because it is the more familiar term. Neither shortcut is good procurement practice.
Why low-boron filter media matters in electronics and semiconductor air filtration
This point gets overlooked all the time.
In semiconductor and microelectronics environments, the problem is not just particles. It is also molecular release from the filter itself. Camfil says its semiconductor prefilter range is designed to protect HEPA filters without releasing boron, and its HEPA/ULPA products for semiconductor use are designed for very low outgassing. AAF makes the same point from the microelectronics side, stating that boron-free and chemically inert filter media limit off-gassing that can damage microelectronics. Hollingsworth & Vose also describes its low-boron microglass media as designed for semiconductor fab cleanrooms.
Why boron is a real issue
In semiconductor manufacturing, trace contamination is not theoretical. Airborne molecular contamination and chemical residues can affect wafer surfaces, optics, and process tools. Camfil explicitly notes that undesired reactions involving acids, bases, organics, refractories, and dopants can create defects and reduce equipment efficiency. Boron is a dopant element in semiconductor processing, so uncontrolled boron release from materials in the air path is exactly the kind of thing fabs try to avoid.
That is why buyers in electronics often ask for:
•low-boron media
•low-outgassing adhesives and components
•ULPA or high-end HEPA/ULPA modules built specifically for microelectronics
•cleanroom filters matched to lithography, etch, thin film, or inspection tool environments
Low boron is not a universal requirement
This is important. Low-boron filter media is not necessary in every cleanroom.
It becomes important when:
the process is semiconductor or microelectronics
the buyer is managing AMC and material outgassing
the fab or toolmaker has tight specs on chemical cleanliness
the cost of trace contamination is extremely high
For many pharmaceutical, hospital, or standard industrial cleanrooms, this requirement may never appear in the RFQ.
Common sourcing mistakes
1. Comparing HEPA at 0.3 μm to ULPA at 0.12 μm without checking the standard
That is an apples-to-oranges comparison unless you confirm the stated test basis and class. EN 1822 / ISO 29463 classification is built around MPPS.
2. Specifying ULPA because it "sounds safer"
Higher efficiency is not a free upgrade if the system cannot absorb the Pressure Drop or if the process does not need it.
3. Ignoring outgassing and chemical cleanliness in electronics
In semiconductor applications, the filter's own materials matter. Low-boron and low-outgassing construction can be as important as the efficiency class.
4. Writing "ULPA filter" in the RFQ without a class
A serious RFQ should say U15, U16, or U17, plus the test standard and size.