PTFE Membrane vs Standard Felt Filter Bags: When Is the Upgrade Worth It?

Record the exact media, fiber, scrim, weight, finish, and membrane identity for each candidate. For the membrane candidate, add membrane side, lamination, and substrate, plus the bonding or production declaration and finished-bag revision. For the unlaminated felt candidate, record fiber blend, scrim, needling or forming description, basis weight, surface finish or treatment, seam, thread, end construction, and finished-bag revision. BWF’s page describes its own industrial filter media as layered fiber coatings and supporting scrim, with ePTFE membrane or treatment depending on application. [5] That manufacturer-specific description illustrates why the full construction record matters; it is not a universal felt or membrane definition. Preserve supplier, product family, grade, part number, drawing, lot or production reference, declaration date, and approved substitutions. Treat PTFE fiber felt, PTFE treatment, PTFE coating, ePTFE membrane laminate, and PTFE seam tape as different descriptions. Treat standard felt as incomplete until its actual construction is traceable. A photograph, texture, color, stiffness, price, or old invoice cannot supply the missing identity.

Donaldson’s Tetratex brochure describes its named ePTFE membrane media as laminated to supporting substrates and presents surface filtration rather than traditional depth filtration for that product family. [3] Gore’s membrane-versus-non-membrane page likewise describes filtering at the surface and dust penetration behavior for GORE membrane constructions and its illustrated non-membrane comparison. [4] These are original manufacturer statements about named technologies, not proof that every membrane laminate shares the same membrane, substrate, pore structure, lamination, durability, pressure behavior, or outcome. They also do not define every unlaminated felt. Use the sources to request exact candidate documentation. Do not copy their diagrams or performance claims into an unidentified product comparison, and do not infer membrane presence from a visually smooth surface.

Document dust size, shape, loading, and stickiness, together with composition sources, fines behavior, cohesiveness, abrasiveness, oil or liquid association, cake observations, process variation, and representative sampling limits. Record gas temperature, moisture, dew point, chemistry, and oxygen across normal, startup, shutdown, peak, upset, washdown, steam, ingress, and seasonal states. Include duration, frequency, measurement location, method, instrument status, and time-aligned process events. A membrane claim developed for one dust and gas condition is not transferable to a sticky, abrasive, condensing, chemically different, or differently loaded duty without relevant evidence. A felt history from another compartment is not automatically a baseline. Unknown chemistry, dust variation, dew-point relationship, or upset exposure remains HOLD. No candidate advances because it has one apparently favorable temperature or chemical-resistance entry.

Capture collector make, model, asset, compartment, revision, airflow direction, filter dimensions, tube sheet, cage or support, spacing, inlet distribution, seal, venturi or blowpipe relationship where applicable, and installed quantity. Record collector geometry, support, cleaning method, and settings history without prescribing settings. Add cleaning controller events, valve or motion maintenance, compressed-air quality where applicable, differential-pressure basis, process rate, and changes during the comparison period. EPA’s baghouse chapter distinguishes shaker, reverse-air, and pulse-jet arrangements and connects cleaning and filter construction to system design. [7] It does not approve a membrane or felt candidate. Preserve installation, storage, packaging, and handling record, including cage condition, protected staging, bag orientation only from the exact manufacturer document, installer, date, and exceptions.

Write the buyer-defined filtration, exposure, product-quality, and economic boundaries first. For filtration or emissions, identify the governing requirement, sampling location, method, instrument, units, averaging basis, process state, conditioning, detection limit, and responsible authority. For pressure and cleaning, align airflow, production, dust loading, collector configuration, filter seasoning, cleaning control, measurement points, time window, and maintenance state. EPA lists outlet particulate concentration or leak indication, differential pressure, inlet temperature, flow, cleaning operation, and fan current as indicators, while warning that conditions affect fabric-filter performance. [2] A bench efficiency, permeability, residual pressure, or cleanability value becomes comparable only when the named finished constructions, method, dust, conditioning, face velocity, loading state, cleaning sequence, and reporting basis align. Otherwise preserve both numbers separately and mark the outcome non-comparable.

Assemble differential-pressure and emissions records for the same operating boundary, then add gas flow, production, fan control, cleaning activity, downtime, maintenance, visible or instrumented events, and process changes. Separate baseline, startup, seasoning, steady operation, abnormal events, and end-of-service periods. A change after installation is temporal evidence, not proof that media construction caused it. Changes in dust, airflow, leaks, seals, cages, support, cleaning, ducting, hopper condition, production, instruments, or maintenance can affect the same records. Define comparison cohorts before selecting dates, and retain excluded periods with reasons. When candidate populations, collectors, methods, or duty differ, do not normalize them into an unsupported percent improvement. Keep efficiency, emissions, differential pressure, cleaning response, life, downtime, energy, cost, ROI, and compatibility as HOLD until the evidence is comparable, attributable, and site/OEM reviewed.

Parker’s BHA Preveil instruction sheet tells users of that named ePTFE membrane product to protect bags during staging and installation and to inspect reused cages for roughness and burr-related fabric risk. [6] Keep those instructions confined to that product and the buyer’s approved procedure; they do not establish a universal membrane-side orientation, installation method, cage acceptance, or cleaning setting. For both candidates, map holes, tears, flex marks, abrasion, seam changes, scrim exposure, surface change, blistering, separation, delamination, deposits, and handling observations to exact positions and populations. Preserve unused controls and apparently unaffected filters where the site program permits. Damage appearance does not identify root cause or prove that membrane, felt, installation, cleaning, chemistry, temperature, support, or handling is responsible.

Build failure, damage, service-life, and changeout history for each traceable cohort. Record installation and removal dates, quantity, locations, duty, current status, removal reason, condition evidence, process changes, maintenance, cleaning, inspections, event investigations, and whether filters remain in service. A still-running cohort has not completed service life; a removed cohort may have been changed for reasons unrelated to media condition. Supplier life estimates remain assumptions unless their basis matches the project. Do not compare the longest membrane run with the shortest felt run, combine different collectors, or omit failed units. Use an approved population and time boundary. If failure cause is unresolved, show it as unresolved rather than charging it to one construction. The lower purchase price does not prove more changeouts; the higher purchase price does not prove fewer.

Define one buyer-selected period and quantity, then enter purchase and delivery, changeout labor, planned and unplanned downtime, attributable fan and compressed-air energy difference, service-life and changeout history, disposal and cleanup, and documented emissions or quality-event cost. EPA treats replacement, labor, utilities, maintenance, and dust disposal as distinct cost categories and says local current inputs should suit the specific system. [7] DOE’s fan sourcebook explains that operating point follows the fan and system resistance relationship, reinforcing that energy attribution requires aligned system conditions and measurements rather than a media label. [8] Use buyer records rather than either source’s examples. An event cost is included only after the buyer’s cause and allocation process supports attribution. A higher price or membrane label does not prove value; a lower price or standard felt label does not prove lower total cost. Blank, unknown, non-comparable, excluded, and measured zero are different states. Any material unresolved category keeps the economic result on HOLD; no partial subtotal becomes ROI, payback, savings, or a ranking.

Upgrade is a purchasing hypothesis, not a result. The buyer-appointed authority should record the collector and duty, candidate constructions, evidence sources and revisions, comparable methods, exclusions, assumptions, unresolved causes, site and OEM review, hazard-review owner, acceptance criteria, monitoring plan, economic boundary, and signed disposition. Use the active filter-media selection guide when the underlying fiber, scrim, temperature, moisture, chemistry, or mechanical duty is still unresolved. Use the active total-cost comparison guide to keep cost categories and holds visible. A conditional candidate is not approved performance, fit, compatibility, savings, compliance, or return to service. If one material field, outcome method, causal link, or full-period cost category remains material and unresolved, record HOLD and the owner of the next evidence action.