When planning a compressed air filter differential pressure monitoring system, factory teams often ask whether a sensor can tell them exactly when to replace an element, how to connect it to an existing PLC, and what acceptance tests should cover. Differential pressure (DP) measures resistance to flow across the filter. A rising trend under comparable conditions can point to premature blockage. It cannot, by itself, prove air purity or the integrity of the element. This guide turns that distinction into a practical installation specification for a Thai manufacturing site.
Define the decision before selecting a differential pressure sensor
Start with an equipment register: filter tag, housing and element model, location in the air system, upstream and downstream isolation valves, normal and maximum flow, operating pressure, downstream users, replacement history and required air quality. A filter on a main header and a final filter immediately before a product-contact process have different consequences when pressure falls. An alarm cannot be useful until the owner and required action are clear.
Separate three objectives. First, detect a rise in resistance at comparable flow. Second, retain DP, flow and operating-state history so maintenance can compare conditions before and after replacement. Third, inspect before insufficient pressure affects downstream equipment. Air-purity verification is a separate test plan. ISO 8573-1:2010 defines purity classes for particles, water and oil; it does not translate a DP reading into a purity class. ISO lists a future revision as under development, while the 2010 edition remains the published version as of November 2026.
Parker warns that a DP indicator is not a stand-alone air-quality or replacement indicator. DP varies with flow, and damaged or torn media may allow air through with deceptively low resistance. A low reading therefore does not mean “safe to keep indefinitely.” Use the element manufacturer’s replacement guidance, actual operating conditions, physical checks and any required downstream quality tests alongside the DP trend. A coloured replacement zone on one product is specific to that model, not a universal threshold.
| Question | What DP can tell you | What else to verify |
|---|---|---|
| Is flow resistance rising? | A trend at comparable flow | Flow, upstream pressure, bypass state |
| Is replacement worth planning? | Evidence of possible early blockage | Model-specific schedule, physical inspection, history |
| Does air meet a purity class? | DP alone cannot establish this | Point-of-use particle, water and oil tests |
| Is downstream pressure sufficient? | One component of pressure loss | Actual downstream pressure and other system losses |

Select and install the compressed air filter DP measurement
Choose pressure taps before the sensor range
DP is P1 minus P2: pressure just upstream minus pressure just downstream of the filter. If the taps are far apart, pipe bends, valves and fittings become part of the reading. Check existing tapping points, isolation for maintenance, condensate accumulation, flow direction and safe access on site. Any retrofit to a pressurised line needs an approved isolation, depressurisation and leak-test plan.
Options include an integrated DP transmitter, a visual indicator and two pressure transmitters with a PLC subtraction. A direct DP transmitter gives one measured difference; a two-transmitter design also provides upstream and downstream pressure but adds calibration mismatch and time-alignment error. Check maximum line and common-mode pressure, overpressure, wetted materials, temperature, range, accuracy, response, enclosure rating, power and signal type against the selected part number. A “low DP” device is not automatically rated for direct connection to a high-pressure plant line. The SMC PSE550 instruction manual, for example, states its own ranges and electrical outputs; those specifications must not be applied to other models.
An existing coloured mechanical indicator may support rounds but may not output continuous data to a PLC. For 4–20 mA, specify wiring distance and open-circuit detection; for IO-Link, master compatibility; for Modbus, registers and communications-failure behaviour. First specify the decisions and the panel constraints, then select the device.
Record flow and operating state with DP
The same clean element can show different DP at different flows. Comparing a quiet night shift with full production creates false alarms. At minimum, record DP, timestamp, upstream pressure and compressor or line running state. Add flow where practical. If a flow meter already exists, use it; if a new one is needed, consult our compressed air flow meter selection guide. Without flow data, compare within documented production modes, shifts or numbers of running lines and mark this as an approximation. Do not hard-code a flow-normalisation formula before seeing actual site data.
An abrupt rise may be caused by a loaded element, condensate, a partially closed valve, a blocked impulse tube, sensor drift or a flow surge. An abrupt fall may result from lower flow, an open bypass, a disconnected sensor or torn media. The alarm should display possible checks in order, not declare one cause automatically. Dew point is another independent indicator: if moisture is a concern, see our compressed air dew-point monitoring guide. Putting DP, drain and dew-point data on one screen does not make them equivalent quality tests.
Connect the PLC, historian and maintenance workflow
A practical chain is pressure taps → DP or two pressure sensors → existing PLC/remote I/O → local historian or gateway → maintenance screen and notification. Where product lots matter, link the time-stamped event to MES records. A plant-local system is enough for many sites; cloud integration is a choice driven by the operating need. The monitoring addition must not disrupt existing machine controls. Plan for communication loss, locally retained records where needed, and an explicit data-quality flag on recovery.
Store the raw value together with units, range, calibration, sensor-health flag, timestamp and filter ID. The screen should show the current DP and flow band, last replacement, manufacturer guidance, baseline period and conditions under which an alarm is valid. A replacement record should include person, time, old and new element part numbers, reason, observed condition and stable post-replacement DP. That record becomes the candidate baseline only after approval.
Separate alerts into measurement failure, inspection candidate and downstream process impact. A maintenance threshold is not a safety interlock. Automatic machine stopping requires its own risk assessment, control design and authority. Do not silently turn a dashboard setting into a protective function.

Decide when to inspect or replace an air filter
There is no universal DP value for replacement. Housing, media, rated flow, pressure, contamination, application and the specific indicator all matter. The colour bands in the Parker Balston C02-2377 manual apply to that model. Use the selected filter manufacturer’s documentation and warranty requirements at part-number level.
With a clean element, record several representative shifts across normal flow bands. Extend the baseline period if production is atypical. An inspection candidate can be defined as a sustained rise from the approved baseline within a comparable flow band, checked against model-specific limits. Persistence time and valid flow bands should be tuned during commissioning against both false alarms and missed events. No threshold quoted in a general article can replace that work.
Keep scheduled replacement and quality requirements visible even when DP is low. Torn media can make a low reading misleading. An unexpectedly low DP is itself a condition to investigate: verify flow, bypass, sensor connection and media condition. Use the following response map as a starting point.
| Event | First checks | Action |
|---|---|---|
| Gradual rise at comparable flow | Replacement date, flow, drain, upstream conditions | Open a planned inspection against model guidance |
| Sudden rise | Valves, condensate, taps, process load | Check sensor and hardware locally |
| Sudden fall | Bypass, wire, media, flow | Do not mark healthy without investigation |
| Stable DP but air-quality failure | Particle, water and oil measurements at the use point | Start a separate quality investigation |
| Bad sensor or communication | Raw value, supply, wiring, clock | Flag data invalid and notify maintenance |
Preserve the reason for replacement, not merely the date. Record the removed element’s condition, drain issues, oil carry-over, upstream pipe work, flow during the event and relevant quality results. If DP rises sooner than expected, investigate contamination sources and upstream treatment before simply shortening the replacement interval. If DP remains high after replacement, inspect valves, pipework and measurement connections.
On a shift-based Thai site, an actionable message needs the equipment ID, location, measured value, comparison flow band, duration, first checks and contact. Thai-language shop-floor screens and consistent equipment IDs across Thai, Japanese and English maintenance records reduce handover errors. Log who acknowledges or changes an alarm and why.
Write an RFP that vendors can quote consistently
“Visualise filter DP” leaves too much unspecified. Include the process diagram, number and type of filters, existing tapping points, pressure and flow conditions, required stoppage window, sensor requirements, signal and tag list, historian, alert ownership, replacement workflow, FAT/SAT and training. Mark items that require a site survey as provisional. Separate sensor prices from pipework, PLC expansion, cabinet modification, gateway, software, commissioning, spares and recurring support.
| RFP area | Minimum content | Evidence to request |
|---|---|---|
| Scope | Filter IDs, process diagram, taps, flow and pressure | Site survey and revised P&ID |
| Measurement | Range, pressure rating, accuracy, calibration, environment | Model-specific datasheets |
| Construction | Isolation, depressurisation, leak test, cabling | Method statement and drawings |
| Data/control | I/O, signals, tags, clock, failure handling | I/O list and logic description |
| Maintenance | Baselines, alerts, replacement records, ownership | Screen mock-ups and workflow |
| Acceptance | FAT, SAT, training, handover | Test forms and manuals |
Pilot one line that matters, has recorded operating conditions and has a feasible shutdown window. A critical line that cannot be stopped may be a poor first installation. Standardise the data model and tag naming before rollout, but keep thresholds specific to filter and operating condition.
For a business case, separate benefits that can actually be measured: avoided excessive pressure loss, fewer pressure-related interruptions, less manual checking and faster quality investigations. Compressor energy effects depend on pressure control and the whole air system. A generic manufacturer’s percentage is not a guaranteed saving for this plant. Compare like-for-like pre- and post-installation measurements and document assumptions for avoided downtime.
Confirm equipment boundaries before requesting a quotation
During the survey, verify that drawing-based pressure taps actually exist and are usable, which valves isolate the filter for element changes, and whether the new sensor will obstruct removal of the bowl or cartridge. Older pipework may differ from the drawings. If new taps are needed, the plant must decide the shutdown window, how other lines receive air during construction and how restart leaks are checked. Include access platforms, cramped work areas and the cable path to the cabinet in the estimate. Ask vendors to itemise tap count, pipe materials and panel changes rather than quote “installation” as one unexplained sum.
Draw the system boundary too. Is the PLC merely calculating DP, is a monitoring server storing history and issuing alerts, or does MES also map events to lots? For each connection, specify what remains available when communication fails. Screen latency, aggregation interval and notification registration belong in the contract; otherwise a vendor may deliver the sensor while treating alerts as separate work. Assign costs and owners for Thai-language screens, training of local maintenance staff and the storage location of a spare transmitter.
Separate measurable and less tangible returns
Lower pressure loss may affect compressor load, but evaluate it alongside pressure setpoints, sequencing, leaks and changes in demand. Compare DP, downstream pressure, flow and electricity under matching production conditions. Better DP alone does not guarantee an immediate reduction in electricity if compressor settings remain unchanged. A plant that raised system pressure to compensate for poor delivery may be able to review that setting after reducing losses, but quantify savings only after the change is measured. Before installation, show the figures as assumptions.
Faster quality investigation also matters even when it is difficult to price. Knowing which filter changed, when and under which operating condition can narrow the process and time window to investigate. DP alone cannot date or prove a contamination incident; combine it with quality samples and equipment inspection. In an investment request, distinguish benefits backed by measured data, operational benefits not yet monetised and questions to evaluate after the pilot. Record how long it takes an operator to identify a filter, inspect it and decide what to do after an alert; compare that with previous rounds. If two filters are in series, measure and label stages separately where justified: a reading on the first cannot diagnose the second, and changing one element must not silently reset the other’s baseline.
FAT and SAT: prove the system works before handover
At FAT, use simulated signals to test zero, normal and high values, open circuits, communication failure, recovery, unit scaling, negative differences, timestamps, alarms and history. Missing data must not be displayed as a valid zero. Test who may change thresholds and whether the audit trail records it.
At SAT, verify actual tap locations and flow direction, safe isolation and leak-test records, wiring, zero and reference checks, and readings across representative running conditions. Trigger a safe simulated alarm and follow it through receipt, acknowledgement, action and reset. Do not deliberately block a live filter to create a test condition. Include Thai screen review, replacement entry, backup/restore instructions and spare-parts handover. A final operational alarm threshold can remain provisional until enough normal-flow data has been approved, provided the approval route is specified.

| Phase | Representative tests | Acceptance evidence |
|---|---|---|
| FAT | Input simulation, scaling, tags, failure, history | Compared input and screen/log records |
| SAT installation | Taps, pipes, leaks, power and wiring | Photos, checklists, leak-test record |
| SAT operation | Baseline flow bands, alerts, replacement event | Real trend, received alert, work record |
| Handover | Access, backup, spares, training | Manuals and signed attendance |
Roll out in stages and preserve the audit trail
First, align maintenance, production and quality on what DP will drive and how air purity is tested separately. Second, survey pressure taps, available shutdown windows, PLC capacity, network and flow data. Third, procure and commission a representative line against the RFP and FAT/SAT plan. Fourth, observe stable conditions and at least one real element-change event before deciding whether to replicate or revise the design. The calendar depends on shutdown access, pipework, procurement and data collection; there is no honest universal rollout duration.
Evaluate the pilot by how many alerts had a confirmed cause, not just by alert count. Record false alerts, missed cases and maintenance response time. A reduction in downtime cannot automatically be attributed to DP monitoring if production mix or air-system configuration changed at the same time.
Retain trend data, alarm context, replacement and calibration records at appropriate resolutions. Agree the retention period with quality and audit owners. Synchronise PLC, gateway and historian clocks so flow, DP and MES lots align. Do not automatically overwrite a baseline when someone enters “replaced”: a half-open valve or drifting zero would become a false normal. Require stable running conditions and authorised approval. Keep old and new threshold values, change reasons and subsequent performance.
For product-sensitive processes, link DP events to lot windows for investigation, but do not automatically declare affected lots defective. The required particle, water and oil tests remain under the quality team’s own plan. Mark the monitored filter and the point of use separately on the P&ID; downstream pipework can introduce other losses or contamination after the filter.
A checklist for choosing the pilot line
Avoid selecting the first filter solely because it is critical. Confirm four conditions. First, trace the actual downstream users on the P&ID and at the machine so everyone understands the consequence of a pressure drop. Second, check whether several past element changes are documented; those records provide a useful comparison. Third, establish whether shifts, product mix or line count change air demand and whether that variation can be measured or logged. Fourth, secure a real shutdown window and a maintenance owner who can witness commissioning. Without those conditions, the sensor may work perfectly while the pilot fails to produce evidence for a rollout decision.
During the pilot, review the proportion of alerts for which a cause was confirmed, not only the total number. Ten messages, nine caused by ordinary flow changes, would add work without improving maintenance. One alert that helps identify a failed drain and explain downstream pressure fluctuations may be more valuable. Record false alarms, suspected missed events, investigation time and the response taken for each filter. If downtime falls, check whether production mix or compressor controls changed during the same period before attributing the full improvement to the monitoring project.
Data retention and change control
Saving only instant DP indefinitely makes the events that matter hard to find. Retain operating trends, the data around alarms, replacement records and calibration records at suitable granularity. Set the retention period with the quality and audit teams before reducing or deleting older samples. Specify a common time source for PLC, gateway and server. A clock mismatch can falsely associate a DP event with the wrong flow condition or MES lot.
A replacement entry must not silently reset the baseline. Immediately after maintenance, a partly open valve or a zero-offset sensor can be captured as a new “normal.” Require stable operation and an approval by the maintenance owner. Apply the same discipline to thresholds: keep the previous value, new value, author, reason and the subsequent effect on false alarms or missed events. This lets the site improve the logic while preventing all alerts from being quietly disabled.
Product-sensitive processes need a separate quality boundary
For coating, food packaging or electronics assembly, showing DP events next to quality-test results can speed investigation. They remain different measurements. Normal DP does not establish that particle, water or oil limits are met, while a passing quality sample does not guarantee that filter resistance will stay low. The quality team must define its own measurement location, frequency, method and pass criteria. When DP becomes abnormal, use timestamps to identify potentially affected production lots for assessment; do not automatically mark them defective.
Pressure and purity at the filter outlet can also differ from conditions at a distant use point. Pipe diameter and simultaneous demand affect terminal pressure; condensate or ageing pipework can affect the air delivered after filtration. Show the monitored filter and the actual point of use as separate positions on the drawing. State the boundary of what the DP installation covers in the RFP, so acceptance is assessed against the agreed scope.
Frequent implementation errors
Installing DP without flow context creates nuisance alarms. Using one threshold for every element ignores model and duty. Treating low DP as proof of purity overlooks torn media and bypass. Selecting from range alone ignores line-pressure rating, condensate, electrical interface and local spare availability. Ending SAT when a number appears on the screen leaves disconnection and recovery untested. Finally, sending only an email without an owner and open-action list lets a shift change bury the alarm.
FAQ
Is one compressed air filter differential pressure sensor enough?
One integrated transmitter can measure DP if its connection and pressure ratings suit the installation. Two pressure transmitters provide both P1 and P2, but their calibration and time alignment affect the calculated difference. Either design benefits from flow or operating-state context.
Can an air filter replacement date be decided from DP alone?
No universal rule applies. Use the selected manufacturer’s schedule and operating-hour guidance, physical checks, application and quality requirements. DP is useful for detecting possible premature blockage; torn media may cause a falsely reassuring low reading.
What must be checked when connecting filter DP to a PLC?
Check spare inputs, signal type, supply, wiring, scaling, disconnection detection, units, timestamps and tags. Assess change control for the existing controller. Any automatic stop requires separate control and safety design.
Does normal DP establish an ISO 8573-1 purity class?
No. The published 2010 standard classifies particle, water and oil contamination; DP is a flow-resistance measurement. Test at the relevant point of use by the appropriate method. Do not cite the under-development revision as an enacted standard.
What should the RFP, FAT and SAT include?
Specify equipment and operating conditions, measurement and construction scope, data and response workflow, and explicit acceptance evidence. FAT checks logic with simulated inputs. SAT checks the installed system, real operating baselines, alarms and handover. Approve model-specific thresholds against site data.
Summary
A useful compressed air filter DP monitoring system links a comparable-condition measurement to model-specific maintenance rules, an actionable alert and tested handover. Low DP is not a certificate of clean air or remaining element life. Keep ISO 8573-1 purity requirements and their tests separate. Start with one line, place DP, flow and replacement history on a shared time axis, and expand only after the data supports real decisions.
For a retrofit at a Thai factory, contact TOMAS TECH at the planning stage. An equipment list and current replacement rules are enough to start reviewing pressure taps, PLC constraints and RFP requirements.