Pneumatic Tool Air Demand and Diversity

Pneumatic Tool Air Demandand Diversity

Count the work, preserve the timing, and verify the air at the tool.

1. Engineering answer: demand is a timed service requirement, not an inventory total

Pneumatic-tool demand should be calculated from each tool's documented air consumption at its actual working condition, the time spent in each operating state, and the tools that can operate together. Keep average consumption, simultaneous flow, event duration and required tool-inlet pressure as separate outputs. A correct average cannot establish that a hose, regulator or compressor arrangement will sustain a demanding event. A list of connected tools cannot establish that every tool draws air continuously.

The practical result is a tool duty diversity schedule. It identifies the equipment, its consumption definition, reference-air basis, working pressure, operator or machine sequence, credible overlaps, background loads and evidence gaps. That schedule becomes a set of requirements for the supply network, local preparation, storage and compressor operation. It should not collapse those separate engineering decisions into an unexplained percentage added to a catalogue total.

A workshop can have low average demand and still experience a sustained pressure problem when two grinders work together. Another installation can have a high connected total but little simultaneous use because its documented production sequence prevents overlap. Both observations can be true without contradicting one another. The difference is in chronology and the required productive outcome, not in whether one calculation looks more conservative.

This page does not assign a compressor model, hose size, safe working pressure or operating permission from tool count alone. Its fictional worked cases explain the method and its limits.

2. Define the assessment boundary and the decision before collecting figures

Identify the boundary as one tool connection, a service bay, a workshop branch, a production cell, a temporary construction manifold or the complete compressed-air station. Mark what lies inside it and where air crosses it. Tool demand downstream of an FRL assembly is not the same accounting boundary as compressor delivery upstream of treatment and storage. A shared branch may also supply cleaning air, vacuum ejectors, drains or machines that are absent from the hand-tool inventory.

State which decision the assessment supports. Annual consumption needs time-weighted air volume. A final connection needs the tool's relevant operating flow and pressure. A branch needs simultaneous flow and acceptable pressure loss. Storage needs the net deficit over time and an allowable pressure band. Compressor selection needs the complete repeatable duty, supported delivery, controls and continuity requirements. One number cannot answer all five questions.

Record existing supply arrangements before changing the demand model. A receiver downstream of a meter can temporarily support tool flow that the meter does not show at the same instant. A temporary compressor can feed a manifold through a second connection. An isolation valve can change which users share a supply. An apparently inconsistent balance may therefore be a boundary issue rather than an inaccurate tool specification.

Define the reporting period and production condition. A morning maintenance campaign, normal manufacturing shift and shutdown-tool period are different cases. Preserve which case a figure belongs to. An hourly average from a mixed day should not be labelled the normal production demand of a single operation. Likewise, a future-work scenario should remain distinct from what has actually been observed.

3. Build a tool register around exact identity and productive duty

Give each material consumer a stable identifier and record its exact model or an explicitly unresolved model field. Include tool function, supply outlet, work material, accessory, normal operating condition, pressure requirement and source revision. A pneumatic grinder's accessory and working load can affect the useful operating point; an impact tool may have different rundown, impact and free-running periods. The catalogue family name is not sufficient evidence for every member of that family.

Record installed, connected, available and simultaneously usable counts separately. A spare stored in a cabinet is not a connected load. A connected spare may leak even when it is not used. Several tools available to one operator may be mutually exclusive during manual work, but an automatically cycling machine can continue after the operator moves to another task. The count must follow the physical operating arrangement, not an assumption based on headcount. [S2; N1]

Do not demand unnecessary precision. A clearly bounded preliminary duration can be useful, provided its uncertainty is visible. A fabricated four-decimal duty fraction is not better than a supported range. Group genuinely identical tools only when their operating definition, pressure, reference basis and schedule are also compatible. Retain individual identities when maintenance condition or operation differs.

4. Read air-consumption labels before using the numbers

A consumption value may describe flow while loaded, flow at free speed, an average over a specified operating cycle, a maximum under a stated condition, or volume per operation. These are not interchangeable inputs. Keep the source wording alongside the normalized field. The supplied demand foundation explicitly requires preserving whether manufacturer information is average, maximum or cycle-based.

A current manufacturer example illustrates the distinction: Chicago Pneumatic's CP1117P60 page separately lists air consumption at load and at free speed. The page demonstrates that the two fields can differ; its figures are not adopted as ratings for the fictional tools in this article. The presence of a free-speed value does not prove the same flow during drilling, and a loaded value does not automatically cover a long free-running period. [R2]

A familiar expression such as “average CFM” needs a denominator: average over trigger-on time, a complete fastening operation, a minute with a specified trigger duty, or a work shift? The word average alone cannot answer that question. Request the missing definition before converting it into an on-state rate.

Do not assume that all manufacturers use one common percentage when publishing average consumption. Dividing every chart value by 0.25 would create unsupported on-state values. Conversely, a clearly documented built-in duty factor should not be ignored. Section 21 uses a fictional, explicitly documented factor to demonstrate the arithmetic without declaring that convention universal.

5. Normalize units without changing the gas or operating definition

Put all additive flow values on a common reporting basis before summing them. Useful project fields include original units, normalized units, absolute reference pressure, reference temperature, dry/moist convention, method and measurement boundary.

For a pure unit conversion, 1 ft³/min equals 0.028316846592 m³/min, and 1 L/s equals 0.06 m³/min. These relationships change units only. They do not convert actual compressed volume inside a hose into free-air delivery, correct a tool's consumption for a different inlet pressure, or establish the dry-air basis of a meter. The same numerical unit can represent unlike quantities when its reporting conditions differ.

A compressor may report delivered capacity referred to specified inlet conditions while a tool document uses another reference convention. Reconcile those quantities before assessing surplus or shortfall. If the relevant state remains unknown, retain a basis-unresolved status; a software field should not silently assign standard atmosphere and room temperature to make a comparison possible.

Distinguish gas-state conversion from performance correction. Changing the reporting state of a known gas quantity does not predict the tool's actual behavior at a new operating pressure. Air motors, impact mechanisms, regulators and nozzles respond differently to pressure and load. The corresponding supplier curve or suitable measurement is needed. Do not apply a blanket proportional pressure correction to every tool merely because a simple gas-law expression is available.

Keep temperature ratios absolute and pressure ratios absolute when the selected relationship requires them. A pressure difference across compatible points has a different purpose from a ratio of thermodynamic states. Unit conversion should be reproducible, but mathematical tidiness must not conceal the missing engineering definition.

6. Separate utilization, simultaneity, coincidence and reserve

It lies between zero and one. A utilization of 0.25 does not mean the tool consumes one quarter of its operating airflow whenever the trigger is pressed. It means the defined state occupies one quarter of the chosen time window. Other states may consume air as well.

Simultaneity describes which consumers operate together. For identical active-flow values, a simultaneous count can be useful. For unlike tools, record the actual combination rather than only the count: two large grinders and two small screwdrivers can have the same count but very different demand. A rule limiting active count does not identify the worst combination unless the eligible tools are also known.

To avoid terminology collisions, this page uses coincidence ratio K_c = Q_group,peak / Σq_i,on for the tool-only boundary, with compatible documented on-state reference flows. This ratio is no greater than one only when the listed on-state values cover all relevant states and the same consumer set appears in numerator and denominator. A result above one should prompt a boundary or rating review, not automatic clipping.

Some engineering conventions call the reciprocal a diversity factor. Neither a mean-duty fraction nor a reciprocal factor should be inserted into another formula without its definition. [N1, N2]

Reserve is a separate provision for identified uncertainty, continuity or future duty. It is not diversity with the sign reversed. Record what it covers and where it has been applied. Do not add a margin to every tool, another to every branch, and a final plant percentage without reviewing overlapping reasons. The supplied demand foundation calls for traceable provisions rather than universal factors.

7. Calculate average demand from explicit operating states

For one constant on-state and one constant off-state over a defined window, average reference flow is q_mean = u q_on + (1 − u)q_off. This is a stated two-state model, not a promise that a real tool consumes a constant rate in either state. If off-state consumption is genuinely zero at the chosen boundary, it reduces to u q_on. A connected leaking hose or continuously supplied auxiliary can make the off-state term nonzero.

For several observed or assumed states, calculate V_ref = Σ(q_j Δt_j/60) when q_j is in m³/min and Δt_j in seconds. The mean is V_ref divided by total valid time in minutes. Keep productive running, free running, repeated impacts, purge, standby and isolated periods separate where their consumption differs. Their durations must describe mutually exclusive states for that tool or the arithmetic double-counts time. [N2 extending S1]

Expected total mean demand is the sum of the compatible individual means plus separately identified background means. That statement does not require independent operation. However, the mean does not determine the peak: correlated tools can have the same individual and group averages as uncorrelated tools while generating much larger overlap events. The difference belongs in the concurrency model.

When production records provide operations rather than time, use documented reference volume per operation times the actual operation count. Include retries, rejected fastenings or cleaning cycles when they consume air inside the reporting boundary. Report volume per accepted product separately from volume per attempted operation; the denominators carry different performance meanings.

Do not multiply a measured whole-branch mean by a second utilization factor derived from the same period. The measurement already includes the observed inactive time. A further adjustment is legitimate only when it represents an explicitly different future schedule or scenario, with the transformation retained. The same caution applies to supplier values that already embed duty averaging.

8. Construct the simultaneous event profile without hiding chronology

Represent each constant-rate teaching event by a tool ID, start time, end time and reference flow. Use half-open intervals [start, end): an event ending at 10 seconds and another starting at 10 seconds do not overlap for a positive duration. This project convention prevents an endpoint in a spreadsheet from creating an artificial extra peak. More detailed measured profiles can replace constant-rate events when required.

At each event boundary, calculate Q_group(t) = Q_background(t) + Σq_i(t). Retain the chronological interval table, not only its maximum. Calculate total volume from the interval areas. The table reveals peak magnitude, how long it persists, the time available for recovery and the particular tools responsible. A sorted duration curve is useful for another purpose, but cannot reconstruct this sequence. [S1; N2]

Check overlapping records for the same asset. Two trigger intervals for one physical tool may be duplicate imports, separate states that need reconciliation or a modelling error. The reference implementation rejects overlapping constant-state records for one asset rather than guessing which should win. Multiple genuinely separate tools should carry separate identities even when they share a model.

A tool's pulse width matters. A volume-per-shot value and a firing rate can establish mean consumption, but do not by themselves establish the instantaneous draw of one shot. Obtain a pulse profile, suitable supplier sizing information or measurement for the local connection question. Spreading every shot evenly over the minute may be acceptable for one consumption estimate and misleading for regulator response.

Preserve unobserved possibilities as explicit cases. A schedule showing one day without overlap is evidence for that day, not proof that overlap is impossible. Distinguish observed chronology, a permitted maximum combination, an enforced sequence and an estimated future pattern. Each can be useful, but they carry different authority for equipment sizing.

9. Treat operator constraints and correlated use as engineering evidence

Operator count can constrain a manual-tool group only when the actual work process supports the restriction. Record whether an operator uses one tool at a time, whether tools continue to run unattended, whether multiple crews share a manifold and whether a helper can operate another tool during setup. A staffing assumption can change when production, overtime or maintenance practices change. It therefore needs a revision and an owner.

Mechanical interlocks, controls and approved operating rules can establish stronger concurrency constraints than a casual observation. Even then, consider startup, manual mode, maintenance, retry and recovery cases. A normal production sequence may stagger actions while a recovery routine energizes several consumers together.

A useful statistical illustration shows why independence matters. Ten identical tools each active 20% of a window have a mean active count of two. Under an explicitly independent Bernoulli model, the probability that at least five are active at a randomly sampled instant is about 3.27935%. If all ten are synchronized—either all active or all inactive—with the same 20% individual duty, that probability is 20%. The average is unchanged. These are mathematical teaching models, not measured probabilities for a workshop. [N2]

The general variance of total flow contains both individual variance terms and pairwise covariance terms. Independence allows those covariance terms to vanish; it is an assumption, not a benefit obtained by installing more tools. Shift starts, common workpieces, production takt and shared operator instructions can create correlated usage. A simple independent model can therefore understate clustered demand.

A pointwise probability is not an event-duration probability. It does not say how long a high-demand episode lasts or how likely one is during an entire shift. Do not choose a receiver or critical-service capacity from a percentile without the chronology, consequence and recovery assessment. A probabilistic scenario supplements, rather than replaces, a documented credible operating case.

10. Long bursts need sustained-duty treatment even with low shift utilization

A grinder that runs for several minutes and then remains unused for a long period can have a small shift-average utilization. During those working minutes, its supply requirement is continuous. Atlas Copco's air-requirement guidance specifically distinguishes such prolonged tool use from a simple intermittent characterization and emphasizes assessing simultaneous users.

Separate three time scales: the tool's internal pulses, the operator's continuous working burst and the overall shift pattern. A compressor can cover the shift mean yet deplete storage during a long burst. Conversely, sizing every component from a shift-averaged number can make a tool underperform as soon as real work starts. Duration must remain attached to both demand and the available supply.

Do not infer full-load operating minutes from the time that a compressor motor is running. The package may be unloaded, modulating, stopped between cycles or replenishing storage. Nor does elapsed tool availability establish trigger-on time. Match the denominator to the intended assessment: energized time, air-connected time, actual tool operation and productive shift time are separate quantities.

For repeated long operations, record the start-to-start period and the available off-time. An apparently sufficient receiver may not recover before the next batch. For continuous production, the supply must support the repeatable mass balance, not merely survive one demonstration. Section 23 develops a deliberately simple storage calculation to make this issue visible.

The production team should confirm the real acceptable working pattern. Artificially extending waits to make an undersized supply appear adequate is a production change, not a neutral engineering correction. Any proposed staggering or reduced concurrency needs agreement that required productivity and safe operation remain satisfied.

11. Preserve pressure at the tool while comparing demand

Tool consumption and useful output belong to an operating condition. Record dynamic pressure at the tool inlet during the specified work, not only the regulator's no-flow reading. The supplied workshop and preparation material keeps this distinction explicit. Chicago Pneumatic's air-line guidance likewise identifies dynamic pressure, inlet interface and hose internal diameter as relevant accessory-selection information. [S2, S3; R3]

A low measured flow can represent inadequate supply rather than a low requirement. A starved tool may run slowly or fail to produce the accepted result. Treating that flow as the satisfied design demand would preserve the defect in the next equipment selection. Retain a service-validity flag: did pressure and productive performance meet the specified condition while the flow was recorded?

Compare the complete pressure path at the relevant simultaneous flow. Where a supported regulator outlet pressure is known, downstream passive losses can be subtracted once to screen the endpoint. The regulator's intentional reduction and its flow-related droop are not interchangeable deductions.

Keep different pressure zones separate. A high-pressure tyre or specialist user should not automatically raise every grinding and fastening outlet. Conversely, reducing a branch to suit one tool must not deprive another user of its documented requirement. The demand schedule should identify the actual pressure zone and endpoint requirement for each group, including temporary connections.

Do not issue a universal working pressure through this page. Equipment limits, maximum permitted pressure, normal target and minimum functional pressure have different meanings. Values in worked examples are invented conditions, not setting instructions. Improving a hose or regulator can change both airflow and tool performance; repeat the relevant assessment after a modification rather than assuming the old demand remains exact.

12. Preserve time resolution, measurement coverage and clock alignment

Choose measurement resolution to answer the question. A one-minute average may be useful for shift consumption but can conceal a six-second overlap that determines branch pressure. Fast sampling alone is not sufficient when the sensor, flow meter or data processing responds slowly. Record sampling interval, integration window, instrument response, clock source and any filtering that affects the event interpretation.

For interval-mean flow, multiply each value by its interval duration. Do not use the arithmetic mean of unequal-duration rows. For point samples, state the interpolation or integration model instead of treating them as interval means without justification. A totaliser gives accumulated volume but does not supply the chronology needed to determine a short peak unless suitable interval differences are available.

Missing records are not zero demand. Identify not measured, invalid, saturated, disconnected and known zero as different states. Calculate valid-time coverage and disclose excluded intervals. A mean over valid time should remain labelled as such; it is not automatically the mean over the full requested window. Where defensible bounds exist for the gap, a separate scenario range can be reported.

Synchronize pressure, flow, trigger states and production evidence when their timing is central. A clock offset can make a pressure drop appear to precede or follow the wrong tool event. Retain raw timestamps and any documented correction. A visually aligned graph without a traceable timing basis is not enough to attribute cause.

The observation period must represent the relevant operations, including changeovers, shared-tool campaigns and nonproduction states when they matter. A complete dataset from an unrepresentative shift remains inadequate for another duty. The measurement plan should explain which cases were captured and which remain untested rather than presenting a high sample count as proof of coverage.

13. Report peaks with their duration and averaging window

Publish separate instantaneous-model peaks, sampled peaks, interval-average maxima and rolling-window maxima. Each describes a different quantity. A measured two-second average cannot be relabelled an instantaneous maximum. A maximum from a constant-rate event model is the model's peak, not proof of every internal air pulse. Keeping those labels distinct prevents a numerical comparison from promising more temporal detail than the evidence contains.

For a fixed rolling window, integrate the flow over that window and divide by its duration. Record whether windows slide continuously in the model, advance by the sample interval or use fixed clock bins. Fixed bins can dilute an event that straddles a boundary differently from a rolling calculation. If the distinction affects a design decision, preserve the original chronological data.

Useful outputs include the largest observed interval flow, the longest continuous time above a selected threshold, volume consumed during the identified event, and pressure minimum during that same event. A threshold should have an engineering meaning, such as a supported supply rate or a branch design case. It is not a generic limit copied from another installation.

A high percentile alone does not identify the worst process case. An event occupying less than 5% of the observations may still be routine and important. Reporting a 95th percentile without discussing the excluded tail can hide a repeated tool failure. The permissible degree of unmet service is a project requirement, not a statistical choice made by the dashboard.

When comparing periods, keep the same units, boundary and averaging rules. A lower reported peak after changing from one-second data to one-minute averages is not demonstrated demand reduction. The method changed. Retain both the operational change and the analysis change so their effects can be distinguished.

14. Separate productive, idle, leakage and auxiliary consumption

Identify intended continuous users, idle-tool draw, leaking connectors, preparation losses, drains, cleaning air and other auxiliary uses. They should appear as separate entries where their causes or improvement actions differ. A pressure-maintaining process, automatic drain or vacuum system may legitimately remain active when handheld tools are idle.

Draw the measurement boundary around each baseline. If a branch measurement already includes its leaks and drains, do not add the same estimated terms again. If tool-only consumption is assembled from catalogue rates, those omitted uses may need separate entries. Use an inclusion field that states what a measured or declared rate already contains.

For shutdown or maintenance work, safe isolation can help distinguish groups under an approved procedure, but the worksheet is not authority to disconnect a line, disable a needed purge or change a protective function. Record the responsible person and the permitted state. A “closed” valve on a drawing does not establish that a boundary was physically isolated during a measurement.

A constant background model can be useful for teaching and preliminary work. Real leakage or purge can change with pressure and equipment state. Record whether the baseline was observed under representative conditions or assumed. When the branch pressure is materially changed, reassess the baseline instead of assuming the same rate without evidence.

Keep avoidable demand visible without subtracting uncompleted savings from the present requirement. A proposed leak repair or nozzle change belongs to an improvement scenario until implemented and verified. The current operating case should show existing demand; the improved case should show its separate evidence and remaining uncertainty.

15. Translate tool-group demand into source and storage requirements

Begin with the chronological demand at the selected downstream boundary. Then account for air consumed between that boundary and the compressor, exports, local storage changes and the actual supply contribution. The corresponding source flow can differ instantaneously from the tool flow. Storage discharging during a burst and recharging afterward is the reason; it is not a contradiction in conservation.

For a defined short event, calculate the reference-air deficit by integrating demand minus continuing supply over time, subject to the available recovery between events. A simple event volume divided by time cannot identify the maximum cumulative deficit if the supply and demand change during the interval. The illustrative reference script tracks a cumulative storage debt that returns toward zero when surplus supply is available.

A basic isothermal receiver screen uses V_receiver = V_deficit,ref × p_ref / (p_high − p_low), with compatible pressure units and equal receiver/reference temperatures. The more general dry-gas mass-balance formulation includes the actual temperature and compressibility assumptions.

The pressure band must still protect the downstream functional pressure after the relevant path losses. A mathematically available receiver volume can be unusable if a restrictive valve prevents air from reaching the tools quickly enough. A receiver located on the wrong side of a controlling restriction does not automatically resolve the complaint. Connection, recharge and pressure behavior remain part of the physical design.

For sustained operation, average supply must cover the repeatable consumption and losses on the matched boundary, with suitable control and reserve. No finite receiver can compensate indefinitely for a recurring net air deficit. Nor does a favorable mean alone prove satisfactory compressor operation: package duty limits, control response, treatment, storage and low-demand behavior remain separate selection checks.

16. Account for preparation, branch restrictions and local peaks together

Choose the demand case that actually passes through each component. A common filter sees the combined users downstream of it; individual tool hoses see their own operating conditions. A component downstream of local storage can see a different short-time flow from an upstream component. The diagram, timing and storage location determine the correct case, not a blanket choice of the highest number in the report.

Do not sum component flow ratings in series. The same steady mass flow must pass through the series path, with intermediate pressures determined by the restrictions and controls. A regulator's rated flow at one allowed droop is not a constant flow available at every outlet pressure.

The group peak should include the required tool combination and other simultaneous uses. A shutdown manifold that serves impact tools and a cleaning nozzle may experience its highest demand when the last fastenings are completed and cleaning begins. Separating these tasks into different spreadsheets can conceal the shared event. The actual production or maintenance sequence should resolve it.

A connection replacement can change useful demand as well as pressure loss. Correcting a restriction may allow a previously starved tool to reach its expected operating flow, while reducing the time needed per job. The result may be higher trigger-on flow and lower air per accepted task. Assess both productivity and consumption rather than declaring every increase in measured flow to be a loss.

Keep flow-related maintenance evidence with the schedule. Filter condition, hose damage, coupler bore and regulator performance can explain why the same inventory behaves differently after months of service. A demand study is stronger when it can distinguish a changed work pattern from a changed supply path.

17. Model future production, crews and temporary tools as separate scenarios

Describe committed changes as new tool identities, operating states, event counts, working minutes or concurrency rules. A second shift may increase annual volume without increasing the normal simultaneous flow. A second crew in the same shift may increase concurrency substantially. A faster production cycle can increase average consumption while also shortening the recovery time between peaks. “Twenty percent growth” does not reveal which change is intended.

Keep current, committed, probable and possible scenarios separate. Identify their dates, assumptions and accountable owner. Infrastructure can sometimes be prepared for a future case without purchasing all compressor capacity immediately, but that is a separate equipment and investment decision. The tool schedule should expose the duty difference rather than preselect the commercial solution.

Temporary tools deserve their own consumption and connection records. Maintenance during a production outage may still share supply with critical instrument users or cleaning operations. Equipment hired for a few days can determine an exceptional peak without belonging to the normal annual duty. A normal-production diversity factor should not automatically be carried into the shutdown scenario.

Reserve for reliability is not simply another tool count. An unavailable compressor, isolated branch or dryer service state changes the supply side of the balance. The demand may remain unchanged while the available service is reduced. Keep the demand case and the supply-availability case linked but individually identifiable.

When a scenario is approved, preserve its controlling operating constraints. A forecast that assumes no more than three large tools simultaneously should state how that restriction is maintained and what triggers reassessment. Otherwise the later operator inherits a hidden limitation while the equipment file appears to promise unrestricted operation.

18. Compare efficiency per accepted task without treating air volume as electricity

Tool efficiency assessment should include the productive result: accepted fastening, removed material, drilled hole, finished area or another agreed unit of work. Record air volume per attempt and per accepted output where rejects or retries matter. Reducing trigger time by compromising the work is not an improvement. Neither is reporting a lower flow caused by pressure starvation as evidence of better efficiency.

Calculate reference-air volume from compatible measured intervals or a clearly stated duty model. Converting that volume into station electricity requires separate evidence: package power, control states, treatment loads, pressure and the relationship between the proposed air reduction and compressor response. A constant energy-per-volume factor can be used only as an explicitly supported or illustrative assumption for its stated purpose. [S1; N2]

Separate avoided consumption, estimated avoided electricity and verified savings. A reduced air demand may initially lead to more unloaded operation rather than a proportional reduction in electrical input. Conversely, a properly controlled source may respond differently. The demand schedule defines the air-side opportunity; energy verification belongs to the relevant measurement and baseline workflow.

Review unnecessary free running, open blowing, poor connections and repeated ineffective tool attempts with the production team. These are candidate investigation areas, not proof that every occurrence is waste. Some processes legitimately need purge or a particular sequence. Preserve the required outcome and authorized operating method when evaluating alternatives.

Show the relevant denominator when comparing tools. A higher airflow during work can accompany a shorter task time, while a low instantaneous draw may take longer. The useful comparison uses supported performance and the whole productive cycle, not a single minimum-CFM column selected in isolation.

19. Diagnose demand complaints from synchronized evidence

Start with the failed outcome: tool speed falls, fastening does not meet the process, pressure is low during overlap, the receiver fails to recover, or the compressor cycles excessively. Record when it happens and what changed. A replacement hose, added operator, altered shift pattern or maintenance state can change the result without a defective compressor.

These patterns are hypotheses, not automatic root-cause diagnoses. More than one mechanism can be present. For example, correlated operation can expose a regulator that was adequate for individual tools; a source response issue can occur at the same time as a local coupling restriction. Segment the evidence rather than forcing every complaint into one category.

Do not defeat preparation, overspeed protection, interlocks or approved pressure limits to make a flow test pass. Measuring equipment and temporary connections must be suitable for the task and installed under competent procedures.

20. Accept the schedule against production, service and evidence requirements

Before the assessment is used for procurement, agree what constitutes a representative normal case, credible peak, exceptional event and committed future case. Establish the required endpoint pressure and productive result with the equipment and process owner. The owner of a spreadsheet cannot create these requirements by choosing a convenient pass threshold.

Review every material input's provenance. Supplier declarations, measured data, calculated values and assumptions should be distinguishable. An unresolved consumption convention can invalidate a confident-looking total. A missing time window can invalidate a reported peak. A supplier operating pressure from a different model can invalidate an otherwise correct calculation. Record the specific issue and affected decision.

Freeze a revision of the consumer list, event schedule, normalized basis and design cases. Retain the physical configuration used for measurements and the transformation from raw data to results. Later changes should identify what was added, removed or reinterpreted. Do not overwrite the accepted baseline to make a new result appear consistent.

The final assessment should report at least the compatible connected total; mean consumption on a named period; peak or scenario flow with duration; required endpoint pressure; included background uses; coverage and unresolved evidence; and separate source/storage implications.

A schedule can be complete as a written requirements record while product selection or physical acceptance remains open. Mark those states accurately. The remaining sections provide fictional calculations and editable records that demonstrate this distinction. No worked example establishes a customer requirement, operating limit, measured workshop result or approved compressor selection.

21. Worked example 1: a published average is reduced twice by mistake

All figures in this example are fictional. Six identical fastening tools each have a teaching consumption record of 6 cfm averaged over a specified cycle. A 25% on-time fraction, constant on-state flow and zero off-state flow. These conditions are supplied for the example; they must not be assumed for an actual manufacturer rating.

The implied on-state flow is 6/0.25 = 24 cfm per tool. The proposed workshop schedule gives each tool an actual on-time fraction of 0.35 over the assessment window. Under the same constant-flow assumption, its new mean is 24 × 0.35 = 8.4 cfm. The six-tool mean is 50.4 cfm. A separately identified constant background load of 8 cfm gives a boundary mean of 58.4 cfm, approximately 1.653704 m³/min on the unchanged reference basis.

The three-tool limit is an additional fictional operating constraint. It does not follow from the 35% duty. Six tools at 35% utilization require 126 tool-seconds of operation per minute in aggregate, so a three-tool maximum is mathematically feasible, but feasibility is not evidence that the sequence is implemented. The production schedule must show that the required work fits the permitted pattern.

If the 25% basis had not been documented, the calculation 6/0.25 would be unsupported. The proper result would be an unresolved on-state flow, not a guessed 24 cfm. If free-running and loaded flow differed, separate those states instead of retaining the constant-rate approximation. The example demonstrates an interpretation error; it does not establish that six real fastening tools require these figures. [N2]

22. Worked example 2: equal air volume, very different simultaneous peaks

Consider a fictional 60-second production cycle. Two tools G1 and G2 each consume 0.8 m³/min while active for 30 seconds. Two tools I1 and I2 each consume 0.6 m³/min while active for six seconds. A separate background load consumes 0.2 m³/min throughout the cycle. All flows share one declared reference basis and are assumed constant in their respective states.

In the aligned schedule, both G tools start at zero and stop at 30 seconds; both I tools start at zero and stop at six seconds. In the staggered schedule, G1 runs from 0–30 seconds and G2 from 30–60 seconds, while I1 runs from 0–6 seconds and I2 from 30–36 seconds. The exact boundary convention is [start, end), so the G tools do not overlap for a positive duration at 30 seconds.

Aligned volume per cycle is (3.0 × 6 + 1.8 × 24 + 0.2 × 30)/60 = 1.12 m³. Staggered volume is (1.6 × 12 + 1.0 × 48)/60 = 1.12 m³. Both schedules therefore have a 1.12 m³/min mean, but their model peaks are 3.0 and 1.6 m³/min respectively. Staggering changes concurrency, not total air consumed in this example.

The connected tool-only on-state total is 2.8 m³/min. Excluding the separately counted background, the aligned coincidence ratio is 2.8/2.8 = 1.0; the staggered ratio is 1.4/2.8 = 0.5. Their reciprocal diversity values are 1.0 and 2.0 under the stated convention. Neither ratio equals the individual G-tool utilization of 0.5 or I-tool utilization of 0.1.

For the aligned case, the largest complete 15-second rolling-window average is (3.0 × 6 + 1.8 × 9)/15 = 2.28 m³/min. Its 60-second average is only 1.12 m³/min. Neither represents the 3.0 m³/min six-second model plateau. The reporting window explains the different numbers.

Staggering is not automatically available in a real process. It must preserve output, task dependencies and approved operating behavior. The example compares two permitted fictional schedules; it does not authorize delaying a real tool or altering a machine sequence. Its constant-rate peaks also do not represent unresolved internal impact pulses. [N2]

23. Worked example 3: storage debt and recovery, not peak multiplied by the whole cycle

Use the aligned profile from example 2 and assume a constant 1.4 m³/min reference-air supply actually reaches the protected boundary throughout the event. The rate is a teaching assumption, not a compressor nameplate value. The tools and background together draw 3.0 m³/min for six seconds, 1.8 m³/min for 24 seconds and 0.2 m³/min for the remaining 30 seconds.

During the first interval, the deficit is (3.0 − 1.4) × 6/60 = 0.16 m³. During the next interval, a further (1.8 − 1.4) × 24/60 = 0.16 m³ is required. Maximum cumulative storage debt is therefore 0.32 m³ by 30 seconds. The later theoretical surplus is (1.4 − 0.2) × 30/60 = 0.60 m³, sufficient in this simplified model to restore the 0.32 m³ debt before the next cycle.

For the staggered profile, each six-second high interval produces a 0.02 m³ deficit. The following lower-demand interval has enough theoretical surplus to repay it before the next high interval. The relevant maximum debt is 0.02 m³, not the sum of two isolated deficits that are separated by full recovery.

The geometric results, approximately 213.3 and 13.3 litres, are not vessel recommendations. They omit real thermal changes, component conductance, pressure-dependent supply, control delay, usable installation volume and justified provisions. The chosen lower pressure must separately satisfy the downstream pressure path. A receiver calculation does not establish that a small vessel and its connection can deliver the required instantaneous flow. [S4; N2]

Now shorten the aligned start-to-start cycle to 40 seconds without changing the 30-second G-tool operation or six-second I-tool operation. The background remains 0.2 m³/min. Demand volume becomes (3 × 6 + 1.8 × 24 + 0.2 × 10)/60 = 1.053333 m³ per cycle; mean demand becomes 1.58 m³/min. The assumed supply provides only 1.4 × 40/60 = 0.933333 m³ per cycle. Storage loses a net 0.12 m³ of reference air every cycle. No finite receiver can sustain that repeating deficit indefinitely. A larger vessel delays the symptom but does not balance the duty.

24. Worked example 4: missing observations and pressure starvation change the conclusion

A fictional ten-minute campaign requests 600 seconds of data. Valid interval-mean flow is available for only 420 seconds. The missing interval is explicitly unknown; it is not known zero flow. Every available rate uses the same reference basis and belongs to the same branch.

Valid reference volume totals 5.90 m³. The valid-time mean is 5.90/(420/60) = 0.842857 m³/min, and valid-time coverage is 420/600 = 70%. Dividing the same volume by all ten minutes gives 0.59 m³/min, but that result silently treats the missing period as zero. It is not an observed full-period mean.

Suppose, solely for an additional scenario, other evidence bounds the missing interval's mean between 0.2 and 3.0 m³/min. That adds between 0.6 and 9.0 m³ during the missing three minutes. Total volume would lie between 6.5 and 14.9 m³, and the full-period mean between 0.65 and 1.49 m³/min under those bounds. These are conditional bounds, not recovered measurements. If the bounds are unsupported, the range should not be issued.

The largest observed interval rate is 2.4 m³/min. It is not an established full-campaign maximum because the missing interval could contain a larger event. It also does not describe any subinterval peak hidden by the 30-second average. The report should therefore label it “maximum valid interval mean observed,” not “maximum tool demand.”

During the 2.4 m³/min interval, fictional tool-inlet pressure is 5.5 bar gauge against a documented fictional minimum of 6.0 bar gauge. The nominal service margin is −0.5 bar. That interval does not demonstrate adequate air service. If the tool was starved, its measured consumption may be below the flow it would draw while delivering the required work at the required pressure.

The correct disposition is to investigate the pressure path and repeat the relevant productive-duty measurement, while preserving the incomplete dataset. It is not to resize a compressor from 0.59 m³/min, claim that 2.4 m³/min is a confirmed peak, or convert the deficit into a universal pressure-correction factor. Missing-data quality and failed service are two separate findings. [N2]

25. Three editable records for a usable tool duty diversity schedule

The following records are designed for manual completion in Word and structured capture by a developer. Blank or unknown fields must remain visibly unresolved. Each submitted record should retain its evidence attachments privately and link to them through stable identifiers. Filling every field with an estimate is not equivalent to receiving supplier and measured evidence.

Record A — Tool identity, consumption definition and operating duty

Record B — Chronological events, constraints and diversity scenarios

Record C — Measurement, service acceptance and controlled changes

The supporting JSON template uses separate tool, event, scenario and verification records so one field does not acquire several meanings. Event records reference tool IDs; scenarios reference their events and constraint revisions. Keep source values and derived values distinguishable. A later correction to a catalogue basis should generate a new result version, not erase the original entry.

26. Frequently asked questions

Should every tool's CFM be added together?

Add compatible on-state flows to describe the connected tool-only total, and use that sum for a simultaneous case only when the tools can genuinely operate together. Normal mean consumption needs state durations, while a branch peak needs the actual combination and chronology. A connected total is useful inventory information, not automatically the compressor requirement.

What is the correct diversity factor for a workshop?

There is no universal factor established by the supplied workshop material. Define the factor before using it, then derive the relevant usage or coincidence from operations, controls, observation or a clearly stated scenario. Do not transfer a percentage from a different workshop simply because tool counts match. Document the consequence of the assumed overlap being exceeded.

Can I multiply average catalogue consumption by my utilization again?

Only after establishing what the published average already includes. If it already represents the same duty window, a second reduction double-counts inactive time. If a different future schedule is required, recover the underlying state relationship from documented information or obtain another supported value. A missing catalogue convention should remain unresolved rather than being replaced by an assumed 25% duty.

Is flow at free speed the same as flow under working load?

Not necessarily. Manufacturer data can publish these as different fields, and the actual task may include both states. Use each value for its own documented state and duration. Do not select whichever value is lower to make the compressor calculation fit. A tool's pressure, load, accessory and productive result should remain associated with the data. [R2]

Can operator count prove how many tools run simultaneously?

It can support a documented manual-work constraint, but it is not proof by itself. Automatic actions, unattended running, extra crews and maintenance modes can invalidate a simple one-tool-per-person assumption. Record the actual sequence and exceptions. A physically or operationally enforced limit is different from a pattern that happened to be observed on one day.

Why does my tool slow down when the compressor gauge looks normal?

The tool receives pressure after the regulator, hose, couplings and other restrictions. Simultaneous demand can expose losses that are small when one tool operates alone. Compare dynamic pressure at the relevant points during the failed task, with the tool events synchronized. A normal compressor-room gauge does not establish adequate service at a remote connection.

Can a larger receiver solve a high tool peak?

It can be assessed for a defined transient deficit, provided its pressure band, location, connection and recharge support the duty. It cannot supply a repeating net air deficit indefinitely or correct every downstream restriction. Preserve the event profile and continuing supply; multiplying peak flow by an arbitrary duration is not a complete receiver selection.

Is a one-minute logger adequate for a fastening-tool study?

It may support a particular consumption calculation, but it can hide shorter overlaps and pressure dips. The appropriate recording method depends on the event and the decision. Preserve sensor response and averaging rules as well as sampling interval. A short-event maximum cannot be claimed from data that average the event away.

What should happen when part of the flow record is missing?

Keep the missing period distinct from zero. Report valid time, coverage, integrated valid volume and the mean over that valid time. Do not call it a full-period mean or a verified full-period peak. Separately supported bounds can inform a scenario, but they do not restore the observations. Plan additional measurement when the gap affects the selection decision.

Does less compressed-air consumption mean the same percentage electricity saving?

Not automatically. Compressor controls, operating state, pressure, treatment and the rest of the plant influence the electrical response. First establish the air-side change and its productive outcome. Use separate energy and baseline evidence for a savings claim.