Pneumatic Machine Air Consumption Acceptance

Required asset: Machine Air Acceptance Protocol

R01 decision: Measure consumption per cycle under agreed productive conditions.

R01 release gate: Require a repeatable agreed operating condition and uncertainty statement.

1 Direct answer

Machine air consumption should be accepted from a time-aligned point-of-use measurement made while the named product, recipe, speed, machine state and reject treatment remain inside an agreed operating condition. The result must state its boundary, reference conditions, measurement window, counters, numerator, denominator and uncertainty. A compressor display, catalogue estimate or unsynchronised production total does not establish air per cycle or air per accepted item.

It does not create a universal test duration, production rate, air allowance, acceptance margin, meter accuracy, purge allocation or compressor-sizing factor.

2 Controlled scope

It covers reference-air volume, cycle and product counters, operating states, productive cycles, rejects, aborted cycles, standby, purge, blow off, start-up, allocation, repeatability and uncertainty.

The page does not certify product quality, machine safety, throughput capability, compressor performance, distribution capacity or energy savings. Those decisions require their own boundaries, methods and authorised owners.

4 Acceptance principle

A valid acceptance result links one measured air volume to the production events that occurred during exactly the same interval. It also identifies the non-cycle demand included in that volume. The test condition must be stable enough to repeat and detailed enough for another competent team to reproduce.

Report more than one normalised result when needed. Air per attempted cycle describes actuation burden. Air per accepted item includes the effect of rejects, retries and losses. State-resolved demand explains standby and purge. None should be relabelled as another metric.

5 Measurement boundary

Define the boundary from an agreed pneumatic connection to every user included in the machine result. Identify the flow-meter position, pressure point, treatment equipment, receiver volume, drains, pilot circuits, vacuum ejectors, cabinet cooling, purge, blow off and maintenance connections.

Close or account for branches that do not belong to the test. A downstream leak remains part of the measured machine demand unless the protocol explicitly separates it with supported evidence. Record shared users that cannot be isolated.

6 Operating states

7 Reference flow and volume

State whether the instrument reports mass flow, reference volume, standard volume, free air or actual compressed volume. Preserve its reference pressure, temperature, humidity and standard. A value in L/min or CFM without a basis is incomplete.

For a trace q ref over the accepted window, reference-air volume is the time integral of q ref. If the meter supplies a totaliser, record start and end totals and verify totaliser behaviour during loss of signal or reverse flow.

8 Pressure and dynamic condition

Record pressure at the flow meter and the machine inlet, including the minimum during the governing cycle. Pressure affects actuator force, valve flow, ejector demand, leaks and machine behaviour. A consumption result obtained at an unrepresentative pressure may not transfer to another setting.

Do not raise pressure merely to pass a timing or production test. Reconcile dynamic loss, regulator behaviour and machine requirements through the responsible engineering process.

9 Counter synchronisation

The flow interval and production counters need a common time basis. Preferred evidence uses controller markers recorded with the meter trace or a documented start and stop handshake. Manual readings can be acceptable only when timing uncertainty is evaluated.

Record attempted cycles, completed cycles, accepted items, rejected items, aborted cycles, retries and multi-item cavities where applicable. Confirm whether counters reset, roll over, buffer or update after the physical event.

10 Agreed productive condition

Define the product code, recipe, tooling, enabled stations, cycle rate or speed range, staffing condition, upstream and downstream availability, ambient state and pneumatic setpoints. Record known optional users.

A condition is repeatable when the operating envelope, allowed interruptions, stabilisation, sample length and acceptance calculation are specified before the run. A vague instruction to run normally is not adequate.

11 Productive cycles

A productive cycle is one that meets the protocol's state and sequence definition. For a single-item machine it may correspond to one completed sequence. For a multi-cavity or batch machine, one cycle can create several items, and the item count must be separate.

Use controller events that reflect real completion. Counting commands instead of completed cycles can hide aborts and retries.

12 Accepted items and denominator control

Define accepted item at the agreed quality-disposition point. An item made during the window but inspected later needs traceable reconciliation. If one cycle produces several items, subtract rejected cavities or units rather than rejecting the entire cycle unless that is the production rule.

Air per accepted item uses accepted output as denominator. It therefore carries the air used by rejects, retries, standby and purge when those volumes remain in the numerator. This is useful, but it must not be confused with air per attempted cycle.

13 Rejects rework and aborted cycles

Record rejected items, their cause groups when available and whether the rejected sequence used a full or partial pneumatic cycle. Record rework loops that occur inside the boundary. An aborted cycle can consume substantial air without incrementing the completed-cycle counter.

Do not remove these events silently. Either retain them in the gross accepted-item result or calculate an additional controlled scenario with an explicit exclusion and reason.

14 Standby and idle demand

Standby demand includes leaks and continuous users while the machine is ready but not cycling. Measure it in a defined stable state. Distinguish planned idle, blocked or starved operation and faulted idle because valves, purges and controls may behave differently.

For a gross production-window metric, standby is normally retained. For diagnostic allocation, multiply supported standby flow by the measured standby duration and report it separately.

15 Purge and continuous users

Process purge, cabinet purge, seal air, vortex cooling and instrument air may continue across productive and idle states. Identify whether each user belongs to the machine boundary and whether its flow is constant, controlled or state dependent.

Allocate continuous purge by actual elapsed time, not by cycle count alone. If purge protects product or equipment, reducing it requires process and equipment approval rather than an accounting adjustment.

16 Blow off vacuum and short events

Blow off, air knives, vacuum ejectors and rapid cylinders can create short peaks that a slow meter may under-resolve. The totalised volume can still be useful if the meter response and integration method are suitable. Peak distribution acceptance requires a compatible dynamic measurement.

Record event duration, overlap and control state. Use exact equipment evidence for any model-based estimate and label it separately from the measured total.

17 Start-up shutdown and changeover

Start-up may include receiver filling, actuator homing, purge stabilisation and repeated checks. Shutdown may vent stored volume. Changeover can add cleaning, tool movement and trial cycles.

A steady productive acceptance should exclude these phases only through synchronised markers and explicit reporting. A shift or batch utility result may include them and should state the allocation basis.

18 Leakage and maintenance state

Leaks inside the accepted boundary contribute to the measured result. Record the machine's maintenance state, recent repairs and open leakage findings. A post-repair result is not directly comparable with an earlier run unless conditions are reconciled.

A no-production standby trace helps identify base demand but does not locate every leak. Leak repair evidence and machine acceptance are related records with different purposes.

19 Concurrency and multi-machine boundaries

When one meter serves several machines, synchronise every machine's state and product counter or limit the claim to group consumption. Do not divide group volume equally without evidence.

Coincident peaks matter to distribution performance, while integrated volume matters to utility accounting. A sequence change can reduce peaks without changing air per accepted item.

20 Treatment distribution and point of use

Confirm that treatment, filtration, condensate handling and pressure meet the machine and process requirements throughout the run. Filters, regulators, valves and tubing can create dynamic pressure loss. A meter installation can also introduce restriction.

Record meter pressure and temperature limits, straight-length or flow-conditioning requirements, orientation, contamination sensitivity and maintenance condition from its documentation.

21 Instrument selection

Select a meter for gas, pressure, temperature, flow range, turndown, event dynamics and installation. Record manufacturer, model, serial number, calibration or verification status, range, output scaling and totaliser resolution.

Avoid assigning a universal meter accuracy. Use the actual device specification, calibration evidence and installation contribution in the uncertainty statement.

22 Installation and zero checks

Install the meter at the agreed boundary without creating unacceptable pressure loss. Confirm direction, units, reference basis, configuration and communication scaling. Check zero or no-flow behaviour using the documented method.

Record bypasses, drains and temporary hoses. Photograph or draw the installation where useful, but do not let an image replace the written boundary.

23 Data acquisition and time alignment

Acquire flow, totalised volume, inlet pressure and machine-state markers on a common time base where possible. Record sample rate, averaging, filtering, data gaps and clock offsets.

The analysis window should begin and end on specified events after stabilisation. Retain raw data and the script, spreadsheet or method used to calculate results.

24 Data quality and exclusions

Create rules before the test for meter saturation, missing samples, counter mismatch, power interruption, manual intervention and out-of-envelope speed. Mark excluded intervals and preserve them in the evidence package.

A visually smooth trace is not proof of valid data. Compare integrated trace volume with the meter totaliser and investigate material disagreement.

25 Uncertainty statement

State the measurand, model, input quantities, uncertainty sources, coverage approach and result. Sources can include meter calibration or specification, installation, reference-condition conversion, time alignment, counter reconciliation, repeatability and allocation.

Do not append a generic percentage. The uncertainty must match the reported metric and accepted window. When project rules use expanded uncertainty, state the coverage factor or confidence convention.

26 Repeatability and run length

Choose run length and repetitions that represent expected variability and provide enough accepted items for the purpose. One short stable segment can miss changeover, micro-stops and reject behaviour.

Report each run, mean or other summary, dispersion and acceptance rule. Investigate drift rather than averaging incompatible states.

27 Acceptance calculations

Gross air per attempted cycle equals total measured reference volume divided by attempted cycles. Gross air per accepted item equals that volume divided by accepted items. State-resolved productive air can subtract separately measured standby or purge only when the protocol permits that allocation.

Report numerator and denominator beside every quotient. Never publish air per product without saying whether product means attempted, completed, accepted or saleable output.

28 Worked case A air per accepted item and denominator

Assumptions. A fictional 30.0 min acceptance window records 21,600 L ref inside the agreed machine boundary. The machine attempts 600 cycles and produces 600 items. Quality disposition identifies 570 accepted items and 24 rejected items; six attempted cycles abort before producing an item. The gross volume includes five minutes of planned standby and a continuous purge.

Method. Gross air per attempted cycle is 21,600 divided by 600, or 36.0 L ref/attempt. Gross air per produced item is 21,600 divided by 594, or 36.36 L ref/produced item. Gross air per accepted item is 21,600 divided by 570, or 37.89 L ref/accepted item. The accepted-item denominator excludes rejects and aborted cycles, while their air remains in the numerator.

Result and interpretation. The acceptance result is 37.89 L ref per accepted item for this named condition, before applying the protocol's uncertainty. Reporting 36.0 L per cycle as though it were air per accepted product would understate the accepted-output metric. The denominator must travel with the number.

29 Worked case B standby and purge allocation

Assumptions. In worked case A, stable ready-standby flow is measured separately at 120 L ref/min for the five standby minutes. Continuous purge is 60 L ref/min across the full 30 min. The protocol permits a diagnostic allocation but retains gross volume for acceptance.

Method. Standby volume is 120 times 5, or 600 L ref. Purge volume is 60 times 30, or 1,800 L ref. The residual event-associated volume is 21,600 minus 600 minus 1,800, or 19,200 L ref. Residual event volume per attempted cycle is 19,200 divided by 600, or 32.0 L ref/attempt.

Result and interpretation. Gross accepted-item consumption remains 37.89 L ref/item because standby and purge occurred during the accepted production window. The 32.0 L ref/attempt residual is a diagnostic allocation, not a replacement acceptance value. Purge reduction requires process approval.

30 Worked case C illustrative uncertainty

Assumptions. The case A gross accepted-item result is 37.89 L ref/item. A fictional uncertainty model assigns standard relative components of 2.0 percent for volume measurement, 1.5 percent for run repeatability and 0.5 percent for timing and counter reconciliation. The project uses root-sum-square combination and coverage factor k equals 2.

Method. Combined standard relative uncertainty is the square root of 2.0 squared plus 1.5 squared plus 0.5 squared, or 2.55 percent. Expanded relative uncertainty is 5.10 percent. Expanded absolute uncertainty is 37.89 times 0.0510, or approximately 1.93 L ref/item.

Result and interpretation. The fictional reported result is 37.89 plus or minus 1.93 L ref per accepted item at k equals 2 for the stated model. This is a worked method, not a universal meter allowance or an acceptance margin. The real project must use its actual evidence and reporting convention.

31 Worked case D concurrent machines

Assumptions. A fictional group meter covers two machines and a common purge. Machine A averages 210 L ref/min, machine B averages 180 L ref/min and common purge is 30 L ref/min during a stable interval. A time-aligned trace shows a brief combined peak of 980 L ref/min when blow off overlaps.

Method. Integrated average group demand is 210 plus 180 plus 30, or 420 L ref/min. The common purge must not be counted twice. The measured peak is retained separately from the average and from each machine's per-item result.

Result and interpretation. Group utility accounting may use the integrated 420 L ref/min condition, while distribution assessment examines the 980 L ref/min event. Individual air-per-item claims require individual synchronised counters and a supported allocation; they cannot be created by halving the group volume.

32 Consumption allocation summary

33 Diagnostic method

34 Commissioning and test sequence

Approve the boundary, protocol, operating condition, instruments, counter definitions, uncertainty plan and acceptance rule before testing. Inspect installation, confirm units and reference conditions, check meter zero and verify counter markers.

Stabilise the machine, run the planned states, capture raw data and document deviations. Reconcile volume and counters immediately. Repeat as required, restore temporary connections and archive configuration, traces, calculations and approvals.

35 Acceptance and release decision

Acceptance must name the product, recipe, speed, productive state, pressure condition, measurement boundary, window, attempted cycles, accepted items, rejects, aborts, standby, purge and uncertainty. It must identify the exact metric and acceptance limit source.

Use Pass, Conditional, Fail or Not tested. A repeatable result outside the agreed operating condition is not a valid acceptance result. A low average demand does not close a dynamic pressure or peak-flow issue.

36 Required asset Machine Air Acceptance Protocol

Use this editable protocol to plan, execute and release a machine-air acceptance test. Unknown fields remain open. Do not convert examples into project limits.

Protocol A identity boundary and objective

Protocol B agreed productive condition

Protocol C instruments and synchronisation

Protocol D state and production record

Protocol E calculation and allocation

Protocol F uncertainty and release

37 Filled fictional protocol extract

38 Frequently asked questions

Why is air per accepted item higher than air per cycle

Rejects and aborted cycles can consume air without increasing accepted output. If their volume stays in the numerator while the denominator counts only accepted items, the quotient increases.

Should standby be removed

Only when the protocol requires a separate diagnostic metric and the allocation is supported. Gross production-window consumption normally retains standby that occurred in the window.

How should purge be treated

Record its boundary, state, flow and elapsed duration. Retain it in gross consumption and show a separate allocation when useful. Do not reduce a required purge without process approval.

Can a compressor controller total prove machine consumption

Usually not by itself. The boundary may include other users, and production counters may not be synchronised. Use a defined machine boundary or a supported group method.

What if one cycle produces several items

Count cycles and items separately. Air per accepted item uses accepted units, including cavity-level rejects where the production definition requires them.

Does air per item size the pipe or receiver

No. It is an integrated utility metric. Distribution sizing also needs dynamic pressure, peak flow, event duration and concurrency.