Air Quality Sampling Plan Builder

Air QualitySampling Plan Builder

Turn a process requirement into a sampling plan that identifies the right point, contaminant, method, operating condition and responsible person.

1. Engineering answer: design the evidence before collecting the sample

An air quality sampling plan defines what must be measured, where the sample must come from, how it will be collected and analysed, which operating conditions it represents, and what decision the result will support. It translates a process requirement into identifiable sampling tasks rather than requesting an undefined “compressed-air quality test.”

The plan starts at the point where the air is used. A favourable result at the dryer outlet does not automatically establish the condition after a receiver, distribution branch, flexible hose or final connection. Equally, a dew-point result answers a moisture question; it does not answer a particle, oil or microbiological question.

The builder therefore connects each requirement to a point, operating state, method, scheduled occasion, evidence record and reviewer. Missing information remains visible. An uncollected sample is not a pass, and an invalid sample is not proof that the process air failed. The final output should tell the sampling team what to do and tell the process owner exactly what the evidence can establish.

A useful plan is not the largest possible list of tests. It is a defensible set of observations that covers the required decisions without confusing convenience, repetition or an attractive certificate with representative evidence.

2. The decision this page owns

The builder connects these tasks without reproducing their entire articles.

Use the worksheet for commissioning, periodic verification, a changed production line, maintenance follow-up, a supplier acceptance requirement or a targeted investigation. Keep each purpose explicit because its point selection and operating conditions may differ. A survey intended to locate contamination is not automatically the same campaign as a test intended to release a product-contact utility.

The resource concerns compressed-air systems. It is not an ambient workplace-air assessment, a breathing-air approval or a clinical-gas certification. Those applications need their own requirements and competent specialists. The manually editable worksheets collect decisions; they do not create missing acceptance limits, laboratory capability or permission to operate.

3. Build a requirement record before selecting instruments

Record the process, the route by which air can affect it, and the source of the requirement. A machine specification, agreed purchase requirement or approved internal quality requirement should have an identity and revision. Preserve the actual wording alongside the measurable interpretation. A description such as “clean and dry” is insufficient to determine a sample method or a numerical acceptance decision.

For each contaminant, specify the quantity and units, the applicable limit or classification, the point where it applies, and the required operating conditions. Where a classification is used, record its edition and complete designation. Do not let a partially entered class label populate unrelated limits automatically. Keep a separate requirement for viable contamination or a process-specific chemical where the process owner requires one.

If the owner has not supplied a limit, the builder may create a characterisation plan, but it must label the acceptance field unresolved. Characterisation can describe an observed condition without pretending to establish compliance.

4. Choose points from the air path, not from the easiest valve

Map the actual route from generation through treatment and storage to the users. Assign stable point IDs to physical locations. Include the sample valve, final hose or connection where it changes the boundary. A sketch or photograph can help the technician find a point, but the record should remain identifiable after a photograph is renamed or a temporary tag disappears.

Separate process acceptance points from diagnostic points. The former support a requirement at a defined use boundary; the latter help locate a change between upstream and downstream locations. A central sample can support treatment assessment without replacing a terminal sample required by the process. Conversely, collecting only a terminal sample can identify a problem but leave its origin unresolved.

Document point-grouping decisions. Similar labels or equal pipe sizes alone do not prove that different users share the same contamination exposure. A grouped plan needs a reason why the selected point represents the stated group, plus the conditions that would invalidate that assumption.

5. Define the operating states that samples must represent

A sampling task is a point-and-state observation, not merely a visit to a location. Record the compressor combination, treatment train in service, bypass positions, relevant demand and process state. The supplied material identifies startup, dryer switching, maintenance, ambient conditions and peaks as reasons that one observation may not describe every state.

Select states because they answer the purpose of the campaign. Routine production may be appropriate for a normal-operation requirement. A restart-sensitive process may need a separately defined restart assessment. A suspected switching excursion needs an observation tied to that event rather than an unrelated quiet-period sample. Do not create a universal “worst case” label without explaining what condition is being challenged.

State what must remain stable during each measurement and what must be recorded if it changes. If another compressor starts or a treatment bypass opens, retain the event with the sample. The reviewer can then decide whether the sample remains suitable for its intended use. Do not delete an adverse operating interval merely because it complicates the report.

Also identify the production constraints: when access is permitted, whether a sampling connection affects supply, and which process owner must be present. The scheduled occasion should be feasible without compromising the utility it is intended to assess.

6. Establish frequency and event triggers without inventing a universal calendar

The builder requires an explicit frequency basis rather than supplying “monthly,” “quarterly” or “annually” as an assumed requirement. The plan owner should consider process consequence, operating variability, previous results, treatment condition, configuration changes and the ability of existing monitoring to detect a developing problem. Any selected interval remains a project decision supported by its recorded rationale. [S1, S2; proposed planning workflow]

Separate scheduled verification from event-triggered sampling. A scheduled campaign observes the system at defined intervals. An event task responds to a relevant change: new pipework, a contaminated connection, a treatment fault, a major service intervention, a changed air source or an unexplained result. Completing the next routine visit does not automatically close an earlier event obligation.

Record the due-date convention. Three calendar months and ninety days are different instructions. State the starting event, allowed scheduling window, responsible person and overdue escalation. Where the process only runs seasonally, link the sample to its actual operating opportunity rather than collecting a meaningless sample from an idle line.

Frequency review should preserve history. Record why a programme is increased or reduced, which evidence supported the change and which point groups remain covered. A quieter reporting period is not, by itself, evidence that less verification is appropriate.

7. Create a contaminant-specific method register

Keep the measurement method separate from the purity requirement. A method describes how evidence is collected; a limit describes the condition that evidence must address. The official method scopes below help prevent category mistakes. They are a routing guide, not a substitute for the complete procedure, laboratory instructions or selected equipment manual. [R1–R6]

Record the exact procedure revision accepted for the campaign. A method update does not justify silently changing an existing contractual basis; the reviewer must settle applicability. The plan should identify the laboratory or measurement provider, equipment, approved sampling arrangement, required sample quantity, reporting capability and relevant limitations.

8. Specify the sampling assembly as part of the method

Treat the sample valve, connector, tubing, pressure-control device and collection system as one measurement path. Record the materials and configuration accepted for the target contaminant. Hardware used for ordinary workshop air is not automatically suitable for trace contamination work. Oil from a connection, particles from disturbed fittings or moisture retained in a line can affect what reaches the instrument.

The assembly must not selectively remove the contaminant being measured without that effect being part of the agreed method. A protective filter ahead of a particle counter, for example, cannot be treated as a neutral accessory when the purpose is to measure upstream particles. Similarly, a conditioning arrangement selected for humidity should not automatically be reused for oil or viable sampling.

Keep separate the normal process path and the temporary sampling path. If a terminal hose is part of the required use boundary, bypassing it changes the question. If the objective is to diagnose the hose, paired points before and after it can be planned, with the difference clearly identified.

Installation and disconnection require competent personnel, suitable pressure ratings and the site's isolation arrangements. The worksheet is not an instruction to loosen pressurised fittings or improvise a sampling connection. Record access and isolation readiness before the task is released.

9. Define flushing and stabilisation by evidence, not a memorised duration

Record the method's preparation requirements and the criterion for beginning the actual measurement. Flushing, stabilisation and collection are different stages. Flushing prepares the sampling path; stabilisation demonstrates that the relevant measured condition is sufficiently settled; collection produces the result used for the stated decision. Their durations should not be merged into one unexplained “test time.”

A calculated line volume can help plan gas use or identify a large dead volume, but a stated number of volume exchanges does not prove that adsorbed moisture, residual oil or biological contamination has been removed. The plan should use the procedure and instrument response appropriate to the target contaminant. Avoid inventing one flushing time that applies to every line material and method.

Record the actual preparation, start time, pressure, sample flow and reason for accepting stability. Where an objective criterion is used, keep its source and the supporting trace. If the result is still moving materially at the end of the available access window, retain a qualified or incomplete task rather than choosing the most favourable reading.

Sampling order also matters. A cleaning or disinfection step appropriate to one task can alter another measurement. Agree whether assemblies are dedicated, exchanged or prepared between analyses, and record the order when it can affect interpretation. These are method-planning decisions, not grounds for automatic background subtraction.

10. Control sample flow, duration and volume basis

A reported concentration needs a known sample-volume basis. Record whether the instrument's flow is actual at the sampling condition or already expressed at a reference condition. Preserve pressure and temperature information when the chosen method requires them. Do not apply a second reference conversion to a flow that has already been normalised. [S1; dimensional calculation design]

For constant flow expressed in litres per minute and collection time in minutes:

Sample volume, V = q × t / 1,000, in m³.

For intervals with different verified flows:

V = Σ(qᵢ × Δtᵢ) / 1,000.

Only intervals belonging to the accepted collection period enter that volume. Purge time, an instrument pause and an unmeasured gap are not automatically valid sample volume. Preserve rejected intervals separately.

Where the method supplies a required volume on the same basis, the arithmetic planning time is t = 1,000 × V_required / q. This does not authorise exceeding the instrument's flow range, collection-medium loading, breakthrough limit or approved duration. The sampling provider must confirm that the quantity and flow are appropriate.

Keep these simple volume calculations distinct from compressor FAD or plant demand. A small instrument sample flow does not quantify the entire process-air flow, and a reference-volume conversion does not correct sampling losses or make an unsuitable probe representative.

11. Plan particle measurements around size bands and sampling integrity

Specify the particle bands, reporting units and measurement boundary required by the process. Record whether results are cumulative counts above a threshold or separate counts within bands. These are different data structures. Summing overlapping cumulative channels can count the same particles more than once. Preserve the instrument channel definitions before deriving any comparison. [Proposed data rules based on S1]

The official ISO 8573-4:2019 scope addresses particle size and number concentration and does not separate solid and liquid particle fractions. A particle count therefore should not automatically be described as an identified mass of dust or oil. Condensation or evaporation can also affect the sampled particle population. [R4]

The plan should identify the accepted connection and any pressure-reduction or dilution arrangement, together with the validated correction that the instrument or method requires. Do not add a generic correction because the main line operates at a higher pressure. Store the original count, accepted volume and method-produced result so another reviewer can reproduce the calculation.

Where a direct count-to-volume calculation is appropriate, number concentration = accepted count / accepted sample volume. Preserve the selected band beside the result. A clean-looking total count cannot compensate for missing evidence in a required band, and a number alone cannot establish a purity class without the relevant requirement and method review.

12. Keep water-vapour, pressure-dew-point and liquid-water questions separate

State the moisture quantity actually required. Pressure dew point is a dew-point value associated with the stated pressure. An atmospheric-pressure measurement should not be compared directly with a line-pressure requirement simply because both readings use degrees Celsius. The official humidity method covers water vapour, not every physical state of water. [R3]

For dew-point sampling, pressure changes, condensation in the sample path, leaks, moisture-permeable materials and stagnant gas can distort representativeness. The instrument manufacturer recommends suitable leak-tight sampling materials, short paths and adequate flow, with attention to sample pressure and condensation. Use the selected instrument's documented conditions rather than adopting a universal flow or waiting time. [R7]

The plan should record the sample-cell pressure alongside the process pressure, the applicable temperature conditions and the stable measurement interval. A dryer-panel indication can be retained as operating context without being presented as independent evidence at a distant process connection.

If the question concerns entrained or accumulated liquid water, create a separate observation or test requirement. A single favourable vapour reading cannot describe all liquid-water events at all points. Conversely, a wet sampling assembly may need an integrity investigation before it can support a conclusion about the process. Keep that distinction visible in the task disposition.

13. Separate oil fractions before assembling a total-oil conclusion

“Oil test” is an incomplete instruction. Identify whether the requirement concerns liquid oil, oil aerosol, oil vapour or a total assembled from applicable fractions. The official scope of ISO 8573-2 excludes oil vapour; ISO 8573-5 addresses that separate measurement. The plan must not label an aerosol-only result as a complete total when the agreed requirement needs additional evidence. [R2, R5]

Record the target measurement, collection medium, accepted sample volume, analytical procedure and reporting limit supplied by the laboratory. For total-oil assessment, identify how the required fractions are represented without overlap. Different methods should not be summed if they measure the same fraction or use incompatible locations, units or operating conditions. Any permitted omission of a fraction needs an explicit method or requirement basis, not an empty field treated as zero.

Store exact report qualifiers. A value reported as below a stated limit is not the same as a measured zero. A numerical total based on replacing every qualified result with zero can appear to pass even when the available evidence is too weak to resolve the specification. Section 23 provides an illustrative example of this problem.

Oil-free compressor technology does not, by itself, establish the quality after intake, distribution and final connections. The supplied treatment architecture retains those separate contamination pathways. The sampling plan should follow the required use boundary rather than deleting oil-related tasks solely because of a technology label.

14. Add viable or process-specific testing only with the right specialist basis

Where the process requires viable microbiological assessment, define it as a separate requirement. Do not infer it from particle counts or from the absence of visible water. The published ISO 8573-7 scope concerns sampling and determining viable, colony-forming contamination; the laboratory's method and reporting conditions must be identified for the result to be interpretable. [R6]

The builder records the laboratory-approved collection arrangement, handling instructions, sample identity, volume basis and reporting requirements. It does not prescribe culture conditions, invent universal acceptance counts or assign a “sterile” status. Sampling, handling and interpretation remain with competent personnel under the process-specific procedure.

A report showing no recovered colonies applies to the sample and method used. It is not proof that all organisms are absent, that other biological hazards were assessed or that the complete utility is sterile. A separate process-specific chemical or biological target likewise needs its own defined method and criterion.

Keep the task linked to the actual exposure point. An upstream sample may be useful diagnostically while leaving a required final-connection assessment incomplete. The plan must disclose that gap rather than allowing one completed specialist test to stand in for every location and use.

15. Match reporting capability to the decision

Before sampling, ask whether the measurement can resolve the required limit. Record the instrument range or laboratory reporting capability and its units. Detection limit, quantification limit and reporting limit are not interchangeable labels; retain the terminology and meaning used in the chosen method. A favourable-looking “less than” statement can still be inadequate when its threshold is above the acceptance limit. [S1; proposed decision workflow]

For example, a result reported as less than 0.020 mg/m³ cannot, on that statement alone, establish a requirement of no more than 0.010 mg/m³. The actual value could lie on either side of the requirement. The correct planning response is to settle a suitable method before collection, or to retain an unresolved acceptance decision if the limitation is discovered afterward.

Define how uncertainty affects acceptance before results are available. If the project selects an interval-based rule for an upper limit L, a conservative option is: support acceptance when x + U ≤ L; support failure when x − U > L; otherwise retain an indeterminate decision. Here U is the agreed uncertainty interval half-width. This is an optional worksheet convention, not a universal rule imposed on every method.

An interval rule is not appropriate to every qualified, censored or microbiological result. Such records need their own agreed treatment. Missing uncertainty must not be silently entered as zero, and rounding must not move a borderline value across the decision limit.

16. Plan quality-control observations without confusing them with production samples

Identify the blanks, duplicate measurements, reference checks or other controls required by the selected method. Their purpose is to examine aspects of the sampling and measurement process. Keep their record types distinct from the process samples so that control records do not inflate the count of production points assessed.

A field blank can help investigate contamination introduced by handling, but it should not be subtracted automatically from every result. The applicable method determines whether correction is permitted and how uncertainty and validity are affected. An unacceptable control may qualify an associated batch; the affected records must remain linked to that control.

Duplicates can reveal disagreement, but averaging is not a substitute for investigating the disagreement. Preserve both results, their sample identities and collection conditions. Two observations at different operating states are not interchangeable duplicates merely because they came from the same valve.

Specify who checks controls, when that review occurs and which decision is withheld while it remains open. A report may have arrived on time while its validity is unresolved. Separating administrative completion from technical validity prevents a complete-looking laboratory folder from being misrepresented as complete process evidence.

17. Agree sample handling and responsibilities before the visit

Confirm that the laboratory or measurement provider accepts the planned method, sample quantity, containers or media, handling conditions and delivery arrangements. Record the agreed instructions rather than inserting generic transport temperatures or holding times. Those details depend on the analyte and method, and the supplied references do not establish one universal instruction.

Use a chain-of-custody record where physical samples are transferred. It should connect the point, collection event, sample identifier, collector, handover and laboratory receipt. For electronic measurements, preserve the original file, instrument identity, timestamps and calculation version. A screenshot of a final number is weaker than a record that retains its measurement context.

Assign responsibility for plan approval, site access, safe connection, collection, analysis, technical review and action. One provider may perform several roles, but the roles should remain explicit. The person who books the visit does not automatically own the process-release decision.

Plan the response to missing access, equipment malfunction or delayed laboratory reporting. Record the next action and owner. Replacing a required sample with an unrelated result because the original appointment failed creates a coverage problem, not a completed task.

18. Use separate states for planning, sampling and acceptance

A single green status hides too much. The builder should show whether the plan is ready, the task was collected, the evidence is valid, and the requirement has been evaluated. These states answer different questions.

Each state change needs a date, role and supporting record. A corrected report should retain its relationship to the superseded report. Retesting creates another event linked to the same requirement, not a replacement that erases the first result.

Where one mandatory contaminant remains untested, the summary must show that gap even if other results meet their limits. The builder can organise evidence and highlight unresolved obligations; it cannot sign a release, award accreditation or certify a complete installation merely by calculating a completion percentage.

19. Define the response to an adverse or unusable result

The plan should identify notification, review and containment responsibilities before sampling begins. For an adverse result, retain the original observation and operating context, notify the responsible owner, and follow the approved process response. For an invalid result, identify why it is unusable and whether the affected process decision must remain on hold. Do not use repeated sampling simply to obtain one favourable number. [S1; proposed action structure]

Separate sample-system investigation from process-system investigation. An unsuitable connector, wrong sample volume or failed control points toward a measurement problem. A valid excursion may require examining treatment, bypasses, distribution or terminal components. Both questions can be open at the same time; a suspected sampling error does not automatically justify ignoring a credible process concern.

Record the corrective action, its scope, the intended verification and the authority that closes the issue. A new satisfactory sample applies to its own point and conditions. Whether it supports wider release depends on the requirement and investigation, not simply on its being the latest result.

20. Report coverage against a fixed set of required obligations

Define a sampling obligation as an identified combination of point, analyte or analysis package, operating state and scheduled occasion. The plan must state its counting convention. A laboratory report may cover several analytes, while one collection event may produce several samples.

Freeze the required set for the reporting period and preserve revisions. An approved removal of an obligation needs its reason; an inaccessible point must not disappear from the denominator merely because it lowers the completion percentage. Control samples and duplicates are tracked separately unless the plan explicitly defines them as additional obligations.

Collection completion = collected obligations / required obligations × 100.

Valid-evidence coverage = obligations with valid evidence / required obligations × 100.

Supported acceptance coverage = obligations with a supported meets decision / required obligations × 100.

Acceptance among evaluated valid results is a different percentage with a different denominator. Display both its numerator and denominator. If a denominator is zero, return “not applicable” or “not calculable,” not an automatic 100%.

Missing evidence can be more important than its numerical share. Identify critical uncovered points and states beside the totals. A high overall percentage does not authorise a blanket quality claim when one essential process requirement remains unresolved.

21. Worked example A — sampling completion is not acceptance coverage

All quantities in this example are fictional. A facility defines six process sampling points, two required operating states and four separately counted contaminant obligations at each point-state combination. The issued plan therefore contains 6 × 2 × 4 = 48 required obligations. This is an example counting design, not a prescribed sample number for every site.

At the reporting date, evidence has been collected for 42 obligations. Four of those records are invalid because a documented sampling problem affected them. The remaining 38 have valid evaluated evidence: 35 meet the illustrative requirements and three do not. Six obligations have not been collected.

Reporting only “92.11% passed” hides ten obligations without valid evidence and three known adverse outcomes. The useful summary preserves all categories and identifies whether any gap affects a critical process.

If a resample later replaces an invalid evidence record for evaluation purposes, the original record remains in history. The denominator stays 48 unless the plan itself is revised and approved. Neither the additional laboratory report nor a repeated sample creates an extra successfully covered process obligation.

The example also shows why a grouped analytical package should have a declared counting rule. A report containing four analytes can satisfy four requirements, but it is not four independent field visits. Planning workload, laboratory cost and technical coverage require related but distinct totals.

22. Worked example B — sample volume and particle concentration

Assume an approved instrument arrangement reports sample flow on the same reference basis required for the particle result. The verified flow is 28.3 L/min, and the accepted collection period is 12 minutes. The volume is 28.3 × 12 / 1,000 = 0.3396 m³. If 34 accepted counts belong to one defined, non-overlapping size band, the arithmetic concentration is 34 / 0.3396 = 100.12 particles/m³.

The result does not receive an automatic purity-class assignment. The required band, instrument suitability, complete method and acceptance limit must still match. A count collected during an invalid flow condition cannot be rescued merely by dividing it by a plausible volume.

For comparison, a variable-flow record contains five valid minutes at 20 L/min and five at 30 L/min. Its integrated volume is (20 × 5 + 30 × 5) / 1,000 = 0.250 m³. Using the final 30 L/min for all ten minutes would incorrectly give 0.300 m³. For the same count, that larger denominator would understate concentration.

If an illustrative method-approved target is 0.500 m³ at 25 L/min on the same basis, the arithmetic collection time is 1,000 × 0.500 / 25 = 20 minutes. Equipment and method limits still govern whether that plan is suitable.

Do not add purge minutes to these collection volumes. If two of the twelve minutes were an unmeasured interruption, the original 0.3396 m³ would no longer be supported as an accepted collected volume. Preserve the actual intervals and obtain a defensible volume or retain the result as unresolved.

23. Worked example C — incomplete oil evidence can conceal the decision

Use a fictional upper total-oil requirement of 0.010 mg/m³ at one identified point and operating state. It is an arithmetic example, not a selected purity class or an actual customer specification. Assume the agreed methods cover the required non-overlapping fractions and use a compatible volume basis.

A quantified liquid/aerosol contribution of 0.003 mg/m³ and a quantified vapour contribution of 0.005 mg/m³ sum to 0.008 mg/m³. The numerical difference below the limit is 0.002 mg/m³. That nominal comparison does not resolve uncertainty or sample validity by itself.

Now replace the vapour result with “less than 0.009 mg/m³.” Keeping the stated qualifier, the nominal total lies between 0.003 and less than 0.012 mg/m³, before considering further measurement uncertainty. Replacing the vapour value with zero would display 0.003 and could suggest an unsupported pass. The available bound crosses the requirement, so this worksheet retains an unresolved decision pending the agreed method review or improved evidence.

In a separate uncertainty illustration, use the quantified total x = 0.008 mg/m³ and an agreed expanded interval half-width U = 0.003 mg/m³. The interval is 0.005 to 0.011 mg/m³. Under the optional conservative rule in Section 15, it overlaps the limit and is indeterminate. It is neither automatically acceptable nor automatically a demonstrated failure.

These are two different limitations: a qualified analytical result and an uncertainty interval. Keep their data types distinct. A missing fraction, an inadequate reporting limit and an interval crossing a limit should not all be converted into the same numeric zero.

24. Worked example D — an upstream pass does not release the terminal point

A fictional investigation samples three locations under a documented common operating condition: P01 after central treatment, P02 at a branch connection, and P03 at the required terminal connection including its normal hose. Each valid particle sample represents 0.500 m³ on the agreed basis and the same non-overlapping size band. Counts are 150, 400 and 1,800 respectively.

Against an illustrative screening limit of 1,000 particles/m³, the nominal results at P01 and P02 are below the limit, while P03 is above it. The nominal terminal-to-branch increase is 2,800 particles/m³. Those comparisons remain subject to the agreed validity and uncertainty treatment.

The pattern supports investigating the path between P02 and P03 and checking the local sampling arrangement. It does not, by itself, prove that the hose is the sole cause or that the central treatment has failed. A controlled investigation must establish what changed, and any corrective verification must retain the original terminal boundary.

Suppose a technician obtains an additional favourable reading at a convenient valve before the hose. That new record cannot close the P03 obligation without changing the question. It is diagnostic evidence at another point. Similarly, averaging the three concentrations would not establish that the final connection meets its own requirement.

The practical outcome is a targeted investigation and a retained terminal-point decision, not a facility-wide “two out of three passed” release.

25. Three editable worksheets for a usable plan

The following tables are manual templates. Add rows as required and preserve separate record identifiers. Blank entries mean information remains to be supplied; they are not default approvals. The plan owner should issue the completed definition before collecting samples, then attach collection and review records without overwriting it.

Worksheet A — requirement, point and operating-state definition

Worksheet B — method, collection and custody record

Worksheet C — evidence review and obligation closeout

The issue package should include the point map, active requirement references, agreed laboratory instructions and the task list. A technician should be able to identify the point and collection conditions without relying on an informal telephone explanation. The reviewer should be able to trace a reported decision back to the same issued plan.

26. Ten practical questions and answers

What information is needed to build the first plan?

Provide the intended air use, process requirement, treatment and distribution arrangement, required points, operating states and existing results. Include any laboratory procedure already agreed. Missing limits or access details should become explicit clarification tasks rather than assumed values. The first useful output may be a point-and-requirement register that shows exactly what remains unresolved.

Is testing the compressor-room outlet enough?

Only for a question whose boundary is that outlet, or for a wider claim supported by an explicit representative-sampling rationale. It does not automatically cover receivers, distribution, hoses or final connections. A required point-of-use decision needs evidence appropriate to that point. Keep central diagnostic evidence and terminal acceptance evidence separate.

Does a dew-point result demonstrate complete air quality?

No. It addresses water-vapour condition on its stated pressure and measurement basis. Particle, oil and any additional process requirements need their own evidence. A completed moisture task must not automatically change the status of other contaminants. The report should identify both the completed measurement and the requirements still open.

Can the same sampling kit be used for every contaminant?

Not without checking its suitability for each method. Materials, preparation, pressure handling and collection arrangements can influence the sample. The plan must identify the accepted configuration for each task. A component that protects one instrument may remove the very contamination another test is supposed to measure. [S1; application of the method-specific boundary]

How often should sampling occur?

Use the approved process requirement and a documented assessment of consequence, variability, history and changes. This builder does not invent a universal interval. Keep routine sampling separate from event-triggered tasks, and record the basis for changing the schedule. A test from last month does not automatically answer a question created by maintenance yesterday. [Proposed planning rule]

Does “not detected” mean zero contamination?

No. Preserve the laboratory's statement, applicable detection or reporting capability, units and method. The result describes what the method could establish in that sample. Before acceptance, check whether that capability can resolve the required limit. Never replace a qualified result with zero simply to produce a complete numerical table. [S1; proposed result-handling rule]

Can several terminal points be represented by one sample?

Only under a documented sampling rationale accepted for the intended decision. Identify the group, shared conditions, selected point, exclusions and review triggers. Physical convenience is not a rationale. If the process specifically requires a separate terminal result, a grouped plan cannot remove that obligation without authorised revision. [Proposed point-selection rule]

What happens when a blank or control is unacceptable?

Retain the control result, identify the associated samples and apply the selected method's validity rules. The affected process evidence may remain qualified or invalid. Do not subtract the blank automatically, hide it, or count the report as acceptable merely because the process result is low. Arrange the documented investigation or recollection required by the reviewer. [S1; proposed workflow]

Does an oil-free compressor remove the need for a sampling plan?

No. Compressor technology is only one part of the air path. Water, particles, intake contaminants, distribution and terminal components remain relevant to the actual process requirement. The plan should follow the required contaminants and use boundary, not assume that a technology label establishes all downstream quality.

What does the completed builder actually deliver?

It delivers a traceable plan, task definitions, manual records and coverage calculations. Execution, laboratory analysis, interpretation and process release remain separate activities. For a project review, provide the completed worksheets and supporting point map; scope and testing arrangements must then be confirmed. No sample result, laboratory accreditation or guaranteed site outcome is created by this document.