A compressed-air project begins at the users, not at the compressor sales sheet. Required pressure, air quality, peak flow, leakage and unloaded operation shape the electricity demand. Replacing a machine while leaving an excessive pressure setpoint or poor sequencing can lock in avoidable consumption. IND090 is the cautious catalogue reference, but its current scope and formula must be checked against the actual replacement and a representative demand campaign.
Confirm the IND090 replacement boundary
Draw the system from compressor room to major users, including receivers, dryers, filters and pressure zones. Identify which compressor is removed and which machines continue sharing the header. IND090 should not be presented as a general credit for every leak repair or distribution improvement completed at the same time. Mark pressure zones and metering locations on the schematic so compressor and distribution effects remain separable.
Log electrical power, delivered flow and pressure at intervals that capture shifts, breaks, weekends and high-demand events. A one-week campaign must be shown to represent annual operation. Production counts let reviewers distinguish a more efficient compressor from a quiet week with fewer pneumatic cycles. Synchronise power, flow, pressure and production records to distinguish equipment efficiency from changed pneumatic demand.
Measure pressure and flow as production changes
Record load, unload, modulation and start-stop behaviour for each existing compressor. Long unloaded periods consume substantial power without useful air and can indicate poor sequencing or an oversized unit. The baseline should retain raw logger files, sensor ranges, calibration information and timestamp alignment. Retain unaggregated data for every machine because total-room power can hide inefficient unloading or poor sequencing.
Pressure at the compressor outlet is not necessarily pressure at the critical user. Measure drops across dryers, filters and long mains, then define the minimum service requirement. Lowering pressure may save energy, but it is a separate operational change unless the selected method incorporates it explicitly. Critical-user pressure should define equivalent service; excess discharge pressure must not be embedded as unavoidable demand.
Size compressors around demand rather than nameplate totals
Artificial demand rises when unregulated users consume more air at higher pressure. Quantify stable production demand, intermittent peaks and leakage periods rather than adding nameplate tool flows. Receiver volume and local storage can meet short peaks without oversizing the base compressor. Separate base demand, short peaks and leakage from the receiver response before selecting compressor capacity.
Compare specific power at the relevant pressure and part-load range, including dryer and cooling auxiliaries where required. Variable-speed control is not inherently superior at every duty. A sequenced mix of fixed and variable machines may perform better than forcing one large compressor to cover the entire profile. Compare specific power across the measured profile and include treatment auxiliaries where the official boundary requires them.
Document sequencing and specific power after start-up
Commission upper and lower pressure bands, master-controller priorities, fallback modes, alarms and dryer dew point. Trend a full production cycle after handover. Confirm that a legacy compressor has not silently resumed lead duty, which would invalidate savings attributed to the new sequence. Force and document fallback tests so a controller or communications failure cannot silently restore an inefficient sequence.
Evidence should pair old and new serial numbers, plates, performance data, electrical diagrams, air-system schematic, measurement report, invoices and acceptance record. Photograph logger placement and header connections. Reconcile implementation dates with the periods selected for before-and-after comparison. The acceptance report should reconcile logger channels, calibration certificates and the exact before-and-after campaign dates.
Keep leaks, artificial demand and production effects separate
Leak reduction is valuable but can make the replacement baseline incomparable if performed between measurements. State its date and quantify it separately. Also disclose new users, changed shifts, pressure alterations and production growth so their effects are not assigned to compressor efficiency. Normalise changed production and users explicitly, while reporting leak repairs and pressure reductions as separate interventions.
Apply only the live IND090 formula and evidence rules, and treat the resulting annual final-energy saving as provisional. Validation can revise the representative hours or accepted boundary. The business case should survive without assuming a fixed CAE volume or payment, especially where future air demand is uncertain. The reference codes for this review are IND090. Use low-demand and growth cases to test economics because future compressed-air requirements may differ from the campaign.