Air Demand Profiling

Air Demand Profiling

New completion based on supplied sources.

Measure the demand that the system must serve

Air demand profiling establishes how much air a plant uses, when it uses it and how quickly that use changes. It supports decisions about compressor mix, controls, storage and distribution by connecting measured flow to production events. Installed compressor capacity and a list of tool ratings are useful context, but neither describes the actual demand profile of an operating plant.

The first task is to define the question. A capacity expansion study needs credible coverage of future and current demand. A pressure-interruption investigation needs sufficient time resolution to reveal the relevant event. A baseload investigation needs non-production observations and knowledge of essential consumers. One dataset may support several decisions, but each conclusion must remain within the period, boundary and resolution actually measured.

Define where demand is measured

Mark the flow measurement point on a current system diagram. State whether it measures compressor delivery, flow after treatment, entry to a distribution zone or consumption at a particular machine. Identify air consumed or stored between that point and the consumers of interest. Purge, drains, leakage, temporary connections and receivers can make two legitimate measurement locations report different profiles.

For a zonal study, document boundary valves and cross-connections. A branch supplied from two directions cannot be treated as a single-inlet zone without accounting for both paths. If only one path is measured, state the limitation. A precise reading at an incomplete boundary can be less useful than a less precise but correctly defined system measurement.

State the flow basis

Record whether the instrument reports actual volume at line conditions, a normalized or standard volume, or another defined reference basis. Retain the configured reference pressure and temperature and any relevant moisture basis. Use absolute pressure and absolute temperature where a gas-state conversion requires them. Do not compare unlike flow bases merely because both displays use cubic metres per minute.

Keep free-air delivery terminology tied to its stated test or reference conditions. A compressor's published capacity may not use the same basis as a site meter. Resolve the definitions before comparing demand with available capacity. When the basis cannot be established, report the values separately and identify the clarification required instead of applying an unexplained correction factor.

Select a meter for the expected duty

Consider the expected flow range, minimum detectable flow, pressure, temperature, moisture, oil contamination, installation length and disturbance from valves or bends. Different technologies have different sensitivities and installation requirements. Confirm the actual instrument's limitations rather than assuming a technology name establishes suitability. Record how the chosen location meets the applicable installation instructions.

A meter sized for a large peak may have limited usefulness at very low idle flow. Conversely, a meter selected for normal operation may saturate during peaks. Range exceedance should be flagged in the data, not clipped silently and reported as the true maximum. Where one instrument cannot resolve both extremes, consider a suitable alternative method or a clearly bounded second measurement.

Match logging resolution to the event

Choose a sampling and recording strategy that can reveal the changes relevant to the decision. A long averaging interval may describe shift consumption but hide a short machine-start surge. A very short interval creates more data without benefit if the sensor response and installation cannot resolve that event. Record sensor response, sampling and stored aggregation separately where they differ.

When short peaks matter, retain synchronized pressure and event information. A momentary flow spike may be supplied partly from storage and may not appear identically at the compressor outlet. Avoid interpreting a smoothed upstream trace as proof that downstream demand is smooth. The measurement architecture should follow the physical system and the timescale of the question.

Capture representative production states

Build a calendar of shifts, product families, cleaning, maintenance, starts, shutdowns and non-production periods. Include known high-demand operations and expected combinations of machines. The required duration follows that calendar. A fixed campaign length is not automatically representative if the important production state occurs outside it.

Record interruptions and unusual conditions during the campaign. If a major line is down, the measured profile describes the reduced plant. That data may still be useful, but it should not silently become the basis for full-production compressor selection. Where future demand is included, show it separately as a scenario with explicit assumptions and obtain the process owner's confirmation.

Build a useful event log

Ask operators to record events that explain changes in flow: line starts, batch changes, blow-off use, cleaning, purging, shutdowns and temporary connections. Use timestamps aligned with the logger. A simple reliable event log is often more valuable than an elaborate form completed from memory at the end of the week.

Link events to equipment or zones where possible. If an unexplained step in flow appears, investigate its timing and duration before labelling it leakage or production demand. Preserve uncertainty when the cause is not known. The profile should distinguish observed correlation from a confirmed cause, especially where several operations change at the same time.

Separate persistent demand from variable demand

Identify the demand that continues through each operating state and the additional demand associated with production. A low-flow plateau may represent leakage, essential controls, treatment consumption or a mixture. Establish its components through controlled observation or approved isolation where practical. Do not assume every non-production cubic metre is avoidable waste.

For architecture decisions, describe the duration and variability of demand rather than forcing it into one average. A stable base may support a different machine strategy from rapid variation over the same average volume. The profile should show how long the system spends in relevant demand bands and what pressure service was maintained within them.

Distinguish peaks from sustained capacity needs

A short peak and a sustained high-demand period place different requirements on supply and storage. Record peak magnitude, duration, repetition and the pressure response. A receiver can buffer a finite event but cannot compensate indefinitely for an ongoing supply deficit. Any storage assessment needs the usable pressure range, volume and recharge opportunity as well as peak flow.

Do not select a compressor solely from the single highest recorded sample. Check whether it represents a real event, sensor artefact or abnormal condition and whether it must be served in the design case. Conversely, do not discard a real critical peak because it occupies little time. The consequence of the event matters alongside its statistical frequency.

Calculate volume and average flow correctly

For interval-average flow Q_i recorded over duration Δt_i, total volume on the stated reference basis is the sum of Q_i × Δt_i. Keep time units compatible with the flow unit. Average flow over valid data is total volume divided by total valid duration. Use time weighting when intervals differ; a simple average of unequal-duration samples can misrepresent consumption.

Missing data are not zero demand. Report valid coverage and identify how gaps affect the result. If an interval is excluded because the meter failed, exclude the corresponding duration from the measured average and explain the limitation. Do not annualize a partial period without a supported operating schedule and a clear statement that the annual result is an estimate.

Worked volume example

Consider an illustrative two-hour period with 60 minutes at 8 m³/min, 30 minutes at 12 m³/min and 30 minutes at 4 m³/min, all on the same reference basis. The volumes are 480, 360 and 120 m³. Total volume is 960 m³, and the time-weighted average is 8 m³/min. These invented values demonstrate the arithmetic and are not site measurements.

The average does not describe every design issue. A consumer requiring 12 m³/min during the middle interval may experience a constraint even though the two-hour average is 8 m³/min. Pressure, supply response and storage during that interval must also be examined. The calculation is therefore a useful summary, not a complete compressor-sizing decision.

Use duration curves alongside time series

A flow-duration curve shows how long demand remains above or within chosen levels. It can help compare operating ranges and identify persistent versus occasional loads. Retain the original time series because sorting values removes event order. Two systems can have the same duration curve but different rapid changes, storage requirements and control behaviour.

State the valid period and weighting used to create the curve. Label unusual operating states separately where they would otherwise distort interpretation. Percentiles can summarize the profile, but a chosen percentile is not automatically a design limit. A critical event outside that percentile may still need to be served or managed through an agreed operating strategy.

Account for stored air and transient mass balance

At a boundary containing storage, compressor delivery and downstream use need not match instant by instant. Rising stored pressure indicates accumulation; falling pressure can indicate that storage is supplying part of the demand. A simple comparison of simultaneous inlet and outlet flow must account for that inventory change and any other included uses or losses.

For steady periods with approximately stable storage conditions, balance comparisons become easier. For transient analysis, use an appropriate mass or reference-volume model with stated temperature and pressure assumptions. Do not treat an unexplained difference between two meters as leakage until storage, meter uncertainty, reference bases and timing have been considered. That distinction prevents misleading loss estimates during normal charging and discharge.

Relate demand to compressor architecture

Use the observed profile to evaluate how the available machines operate across demand states. Identify the base duty, trim requirement, minimum stable operating conditions and periods of excess capacity. Review control interactions and pressure performance, not just summed nameplate flow. Actual available capacity can depend on operating pressure, condition and environmental factors.

A variable-speed machine is not automatically the best answer to every variable profile. Its applicable range, efficiency, integration and backup requirements must fit the system. Likewise, adding a fixed-speed unit may worsen low-demand operation if sequencing is not addressed. The profile should support a project-specific comparison with the relevant equipment evidence rather than a technology slogan.

Keep redundancy separate from demand margin

Normal demand, future growth, uncertainty allowance and standby capacity serve different purposes. Show them separately. Adding several overlapping percentage margins can create excessive installed capacity and poor operating behaviour. The process owner should define the service required during an outage and the operating states that may be curtailed if a machine is unavailable.

Where temporary air is part of the continuity plan, include its pressure, quality, connection and mobilization constraints. Do not count a rental option as immediately available without confirmation. A demand profile supports the required duty, while availability and response arrangements remain separate commercial and operational decisions.

Investigate low demand and idle operation

Non-production periods can reveal persistent use that deserves investigation. Compare flow with known essential consumers and treatment states. Approved isolation of a zone may help identify its contribution, but it must not interrupt critical services or create an unsafe condition. Record each valve change and allow the system to stabilize before interpreting the result.

If idle demand falls after repairs, measure how compressor operation changes. Reduced flow can lead to longer unloaded periods rather than a proportional power reduction. Treat airflow recovery and energy savings as separate outcomes. Unsupported savings claims remain excluded from the project conclusion until the corresponding power and operating evidence exist.

Validate the profile before making a selection

Review meter range, reference basis, calibration evidence, clock alignment, valid coverage and production representativeness. Compare obvious events against operator records. Investigate values inconsistent with the physical system. A profile that passes a plotting script but fails these checks should not be used to justify a capital purchase.

Document which decisions the data can support. It may be sufficient to identify an idle-flow problem but insufficient to size a future production expansion. It may capture weekly volume while missing sub-second demand events. Clear limits help the owner commission the right next step and prevent one campaign from being used indefinitely for unrelated decisions.

Profile acceptance matrix

Build a future demand scenario without disguising it as measurement

For planned equipment, collect the specified flow, pressure, duty cycle and operating sequence from the responsible process or equipment source. Identify which new consumers will operate simultaneously with existing peaks. Show the scenario separately from the measured baseline, with each assumption labelled. Do not add all nameplate maxima and then apply an unexplained reduction factor to make the result convenient.

Test more than one plausible operating case where the production plan is uncertain. A scenario can support a provisional capacity range, but it cannot establish the same confidence as observed operation. State what must be confirmed during commissioning and how controls or temporary capacity will manage the uncertainty. This makes future growth explicit without disguising it as recovered site evidence.

Reconcile zone totals with the main meter

When several zones are logged, compare their time-aligned totals with the main boundary over appropriate intervals. Investigate differences through unmetered branches, treatment use, storage changes, reference bases and instrument uncertainty. A mismatch is a diagnostic lead, not automatic proof of leakage between the meters. Check the system topology before assigning the difference to any one zone.

Use stable operating periods for an initial reconciliation where possible. For transients, account for air stored between boundaries and differences in response. Record the residual and its uncertainty rather than forcing totals to agree through an unexplained adjustment. A transparent unresolved balance is more useful than a falsely exact allocation of consumption to production areas.

Distinguish consumption intensity from efficiency

Air volume per unit of product can help compare similar production periods, but its interpretation depends on product mix, quality, rework and supporting operations. A lower ratio may reflect a different product or less cleaning rather than a technical improvement. Define the production denominator and retain its source. Do not use a single ratio to rank unlike processes without a justified basis.

Electrical energy per unit of product adds the station's energy response and may be useful for the wider business decision. Keep air-demand intensity and energy intensity separately reported so the source of a change can be investigated. Neither metric should become a verified savings claim without comparable conditions and appropriate measurements.

Document the decision that follows the profile

The final profile report should say whether the evidence supports a supply-capacity review, a control adjustment study, a local storage investigation, leak work or further measurement. Link that decision to specific operating states and constraints. Avoid ending with a generic recommendation to optimize the system when the measured pattern points to a more definite next step.

Also record what the profile does not settle. It may not establish air quality, pressure-vessel suitability, the condition of an airend or the performance of a particular proposed compressor. Those questions need their own evidence. Preserving these boundaries prevents a useful demand study from being stretched into unsupported approval of the entire project.

Questions from engineering teams

Can connected load be added to size the station

It can establish an initial inventory, but actual simultaneity, duty and service requirements need evaluation. Do not apply an arbitrary diversity factor without a basis. Use measured profiles where available and clearly separate future scenarios from observed demand.

Is the highest recorded flow the required compressor size

Not necessarily. Determine the event's validity, duration and consequence, then evaluate storage, control response and sustained supply. A single sample is not a complete design basis.

Does low weekend flow prove leakage

No. Essential users, treatment and other legitimate loads may remain active. Identify the components of the flow and record the method before classifying avoidable waste.

Can two months of data replace a missing critical event

Long duration does not guarantee representative coverage. If the event never occurred, its requirement remains unverified. Obtain a controlled observation or a justified separate design case.

Enquiry information and next action

Provide the system diagram, production calendar, known critical events, compressor inventory and available flow records. State the flow basis and meter location if known. Reddy Compressors can use these inputs to define an appropriate profile study and identify the evidence required for capacity, controls or storage decisions.

Keep PROJECT ENGINEERED and PROJECT-SPECIFIC status. No sizing or savings conclusion should be drawn from a short, unexplained or unrepresentative snapshot.