Why compressed-air savings need verification
Leak repairs, pressure reduction, a new VSD compressor, a sequencer or a dryer change are often sold on predicted savings. Months later, the electricity bill may show something different, because production rose, the season changed, a machine broke down or the prediction was simply optimistic. Without a defined method, nobody can tell which.
M&V answers a controlled question: what changed, under which operating conditions, by how much, and with what confidence, without compromising pressure, air quality, reliability or production. It is not about building a dashboard. It is about evidence that an engineer, an auditor or a finance team can check.
Define the question and the boundaries
Start with one clear question, for example: “How much compressor energy did the leak-repair programme save per week of normal production?” Then fix the boundaries:
- Equipment boundary: which compressors, dryers, cooling fans, pumps and ventilation are included.
- Time boundary: the baseline period, the implementation date and the verification period, plus any excluded abnormal days.
- Production boundary: the shifts, output levels or product mix needed to compare like with like.
- Pressure point: where pressure is measured and why that point represents the system or critical user.
- Flow basis: whether the meter reports actual volume, free air delivery or a standard/normal reference volume.
If the boundary changes between baseline and verification, for example a dryer is added inside the measured feed, the comparison is no longer valid unless it is adjusted.
What should be measured?
| Variable | Typical instrument | Key points |
|---|---|---|
| Electrical power and energy | Three-phase power analyser or energy meter | True power (kW), not current alone; include power factor and all phases |
| Airflow | Thermal mass or vortex flow meter on the main header | Straight pipe runs, moisture and reference conditions affect accuracy |
| Pressure | Pressure transducers at the compressor and at critical users | Gauge versus absolute; suitable range for good resolution |
| Machine state | Controller data or current logging | Loaded, unloaded, idle and stopped time for each compressor |
| Temperature and dew point | Ambient, inlet and dew-point sensors | Explains seasonal changes and dryer performance |
| Production | Plant records or meters | Output, shifts or running lines to normalise the baseline |
Instrument choice matters. A flow meter running at the bottom of a very wide range gives poor resolution. A display with many decimal places is not necessarily accurate. Record each instrument’s identity, location, range, units, sampling interval and calibration status.
Building a fair baseline
A baseline is not simply last month’s kWh. It is a model of how much energy the system would have used under defined conditions if nothing had changed. To be representative, it should cover every shift pattern, day type and machine combination that materially affects demand, and it should include seasonal effects where they matter. In Telangana and Andhra Pradesh, the difference between a cool January and a 45 °C May can noticeably change compressor and dryer behaviour.
- Log power, flow, pressure and machine states for a period long enough to capture normal variation, often at least one to two full production weeks.
- Mark abnormal periods such as breakdowns, shutdowns, commissioning or a temporary rental compressor, and treat them explicitly rather than blending them in.
- Choose the normalising variable: production output, running hours, delivered air or ambient temperature.
- Check the relationship. A simple ratio works when energy scales cleanly with production; a regression model is better when there is a fixed base load such as leakage.
Key performance indicators that mean something
The most useful single indicator is specific energy: kWh of input per m³ of delivered air, or its equivalent specific power in kW per m³/min at a stated pressure. It links electrical input to useful output and exposes deterioration, poor control and excess pressure.
Worked example (illustrative): a compressor station uses 4,800 kWh in a day and the main flow meter records 26,000 m³ of free air delivered. Specific energy is 4,800 ÷ 26,000 ≈ 0.185 kWh/m³, which is about 11.1 kW per m³/min. If, after a pressure reduction and sequencing change, the same production needs 4,350 kWh for 25,500 m³, specific energy falls to about 0.171 kWh/m³. The verified result separates the effect of using less air from the effect of producing it more efficiently.
| KPI | What it shows |
|---|---|
| Specific energy (kWh/m³) | Overall efficiency of producing delivered air |
| Non-production demand (flow at zero production) | Leakage and open drains or valves |
| Unloaded energy share | Energy wasted running unloaded or idling |
| Pressure stability at critical users | Whether savings have compromised the process |
| Dryer and treatment energy | Purge losses and dryer power |
| Recovered heat used | Heat-recovery performance, in kWh of heat displaced |
As a typical rule, each extra bar of discharge pressure on an oil-injected screw compressor costs about 7% more power, so pressure changes should be logged precisely during any verification.
Data quality: the part most projects skip
- Time synchronisation: power, flow and pressure loggers must share a common clock, or events will not line up.
- Sampling rate: fast events such as load-unload cycles need seconds; daily energy needs minutes. Match the rate to the question.
- Completeness: record gaps and their cause; do not silently fill them.
- Validation: check for flat lines, impossible values, unit errors and gauge-versus-absolute confusion before analysis.
- Raw data: keep the original files unchanged and document every adjustment, so the calculation can be repeated.
- Units: state bar(g) or bar(a), kW or kWh, m³ at which reference conditions, every time.
Common M&V mistakes
- Comparing a busy baseline month with a quiet verification month without normalising for production.
- Measuring current on one phase and calling it power.
- Quoting flow without stating the reference conditions, so free air delivery and normal cubic metres get mixed.
- Leaving a temporary rental compressor or a breakdown period inside the baseline.
- Ignoring the dryer, purge air and cooling fans when the improvement affects them.
- Claiming a saving from lower pressure while a critical machine starts to run short.
Each of these can make a good project look poor or a weak project look excellent. A short written measurement plan, agreed before logging starts, prevents most of them.
Matching the method to the measure
| Improvement | What to verify | Useful period |
|---|---|---|
| Leak repair | Flow and power at zero or low production, before and after | Nights, weekends or shutdown days |
| Pressure reduction | Discharge pressure, power, and pressure at critical users | Full production cycles |
| Sequencing or VSD | Loaded and unloaded time per machine, specific energy | All shift patterns |
| Dryer change | Dryer power, purge flow, dew point | Hot and humid periods as well as cool ones |
| Heat recovery | Heat delivered and fuel or electricity displaced | Periods when the heat is actually used |
Reporting a verified result
A good M&V report states the question, the boundary, the instruments, the baseline model, the adjustments, the verified change and its uncertainty. It also confirms that pressure at critical users, air quality and reliability were maintained. A saving that starves a production line of pressure is not a saving.
Results should be expressed in physical units first, kWh per year, kWh per m³ and m³ of leakage avoided, and only then converted into money using the site’s own tariff. Savings also need to persist: leaks return, pressure setpoints creep back up and machines drift out of sequence. A short repeat measurement after some months shows whether the improvement has held.
How Reddy Compressors helps
Reddy Compressors, based at Tarnaka in Secunderabad and working with industrial compressed air since 2002, carries out air audits, demand and pressure profiling, leak surveys and power and flow logging on systems of any brand. As an independent multi-brand specialist, not an authorised dealer, we can measure and recommend without favouring one manufacturer. We have tie-ups with all major and reputed brands.
We can set the boundary, install logging, build a baseline, carry out or support the improvement, whether leak repair, pressure reduction, sequencing, a VSD or a dryer change, and repeat the measurement to verify the result. Our service teams cover the whole of Telangana and Andhra Pradesh, and we support projects across India where feasible.
What to send us
- A layout or list of compressors, dryers and main pipework with nameplate details.
- The improvement planned or already made, with its date.
- Any existing energy meter, flow meter or controller data.
- Shift pattern, production records and planned shutdowns.
- Required pressure at the most demanding users.
- Electricity tariff structure, if you want results converted into cost.
Frequently asked questions
How do I verify compressed-air energy savings?
Measure power, airflow, pressure and machine states before and after the change, within the same boundary and under comparable production and ambient conditions. Build a baseline that accounts for production and season, exclude abnormal periods, and compare specific energy as well as total kWh. Confirm that pressure at critical users was maintained, then repeat after a few months.
What is specific energy for a compressed-air system?
Specific energy is electrical energy input divided by delivered air, usually in kWh per m³, or its equivalent specific power in kW per m³/min at a stated pressure. It is the most useful indicator of how efficiently a system produces air. It must be compared at the same pressure and reference conditions to be meaningful.
How long should a compressed-air baseline be measured?
Long enough to capture all normal operating patterns: each shift, weekdays and weekends, and the usual machine combinations. For many plants that means at least one to two full production weeks. Where season strongly affects demand or dryer load, a longer baseline or a temperature-adjusted model gives a fairer comparison.
Can I use the electricity bill to measure compressor savings?
Rarely on its own. A site bill includes every load in the plant, and changes in production, weather or other equipment can easily hide or exaggerate a compressed-air saving. A dedicated meter on the compressor feed, ideally with flow and pressure logging, gives a far clearer result that can be normalised and checked.
What flow meter is best for compressed air?
Thermal mass and vortex meters are widely used on compressed-air headers. The right choice depends on pipe size, flow range, moisture, pressure and the reference conditions you need to report. Installation is as important as the meter: follow the required straight pipe lengths and place it after the dryer where moisture could affect readings.
Why did my energy savings disappear after a few months?
Common reasons are new leaks, pressure setpoints raised again, sequencing changed during maintenance, drains failing open or production patterns shifting. Savings in compressed air need ongoing management. A short repeat measurement, a leak-tagging routine and locked controller settings help ensure the verified improvement persists.
Related pages
Calculators and worked examples
These tools give indicative figures for screening, comparison and for preparing an enquiry. They run in your browser and send nothing. No prices or tariffs are assumed: enter your own to see costs. Efficiencies and other figures shown are example values: replace them with your own. Results are not a performance guarantee or a design certificate; final selection is confirmed by a service engineer against the machine data sheet and the site.
Specific power comparison
Compare two compressors on what matters over life: the electrical power needed per unit of air delivered, and what that costs a year.
The worked example below shows the method with fixed numbers. Turn on JavaScript to change the inputs.
Worked example
Inputs
- Machine A: input power (kW): 75
- Machine A: delivered flow (m³/min): 11.5
- Machine B: input power (kW): 80
- Machine B: delivered flow (m³/min): 12.6
- Running hours per year: 6000
- Air actually needed (m³/min): 11
Result
- Machine A specific power: 6.52 kW per m³/min
- Machine B specific power: 6.35 kW per m³/min
- Energy to deliver the air needed: A: 4,30,435 kWh per year
- Energy to deliver the air needed: B: 4,19,048 kWh per year
- Energy difference: 11,387 kWh per year in favour of B
- Cost difference: Enter your electricity tariff (cost per kWh) above
Method. Specific power = input kW ÷ delivered flow (m³/min) at the same pressure and reference conditions. Annual cost = kW × hours × tariff. Use the full-load input power on the same test basis for both machines.
What to take from it. A small gap in specific power is multiplied by thousands of running hours. Compare on the same pressure and reference conditions.
Technical content reviewed by D. Srinivas Reddy, Managing Partner and service engineer, Reddy Compressors. Last updated 17 October 2026. How our engineering content is reviewed.
Baseline regression fit (one variable)
Fit energy = a + b × production to baseline-period data, check the fit, then apply it to a reporting period to get the adjusted baseline and avoided energy.
The worked example below shows the method with fixed numbers. Turn on JavaScript to change the inputs.
Worked example
Inputs
- Baseline energy, one value per period (kWh): 59030, 58120, 60810, 56840, 61440, 61420, 58210, 55120, 58980, 59090, 62150, 58780
- Baseline production for the same periods, same order (units): 19800, 21200, 22600, 20400, 23100, 24500, 18600, 17900, 20800, 22100, 23700, 21500
- Reporting period: production (units): 24000
- Non-routine adjustment, + or − (kWh): 0
- Measured reporting-period energy (kWh): 52400
Result
- Periods used: 12
- Constant a (base load): 39,988 kWh per period
- Slope b: 0.8983 kWh per unit of production
- R²: 0.77
- RMSE: 1,027 kWh per period
- CV(RMSE): 1.7 % (within the 20% screening guide for monthly regression baselines)
- Adjusted baseline for the reporting period: 61,546 kWh
- Avoided energy: 9,146 kWh = 14.9% of the adjusted baseline
- Avoided energy ÷ RMSE: 8.9 more than 2 × RMSE (screening value): clearly larger than the model scatter
Method. Ordinary least squares on n baseline periods: b = Σ(x − x̄)(E − Ē) ÷ Σ(x − x̄)², a = Ē − b × x̄. R² = 1 − SS_res ÷ SS_tot. RMSE = √(SS_res ÷ (n − 2)); CV(RMSE) = RMSE ÷ Ē × 100. Screening guide used by this tool: 20% for monthly regression baselines (the figure commonly cited from ASHRAE Guideline 14; check the edition you report against). The 2 × RMSE test for the saving and the 6-period minimum are also screening values used by this tool. Adjusted baseline = a + b × reporting production + non-routine adjustment; avoided energy = adjusted baseline − measured reporting energy. One driver only: for production plus days or temperature, fit with a spreadsheet and use the adjusted-baseline tool above. Use the same metering boundary in both periods.
What to take from it. Trust the avoided energy only if the fit is good (high R², low CV(RMSE)), the reporting production is inside the baseline range, and the saving is clearly larger than the RMSE.