Who this page is for: maintenance managers, utility engineers and plant heads who have a second compressor marked “standby” and need evidence, rather than hope, that it will start, load and hold pressure at the critical user when the duty machine trips in the middle of a shift.
Key facts
- A standby compressor proves its worth in the first 30–60 seconds after the duty machine trips. That is when receiver storage is draining and the critical user decides whether production stops.
- With this page’s defaults (2,000 litres of storage, a 1 bar (14.5 psig) band from 7.0 to 6.0 bar(g) (101.5 to 87.0 psig), and 8 m³/min or 282.5 cfm of demand), the plant rides through for only 14.8 seconds. A standby taking 30 seconds needs about 4,052 litres.
- Usable free air in storage is about V × (Ptrip − Plow) ÷ 1.013. Each 1,000 litres per 1 bar of usable band gives roughly 0.99 m³ of free air.
- A no-load start proves almost nothing. The test has to load the standby against real plant demand and record pressure at the critical user, not at the compressor house.
- Compressor rooms in Telangana and Andhra Pradesh reach 45–50 °C in summer. A standby that carries the load in December can fall short on capacity or trip on high temperature in May.
- Rotating duty and standby roughly every 500–1,000 running hours keeps both machines exercised and their running hours within a few hundred of each other.
Why installed is not the same as ready
The drawing shows “duty” and “standby”, and everyone feels covered. Then the duty machine trips on high temperature at 3 p.m. in May. Pressure at the critical line falls within seconds, and the standby does not start, starts but does not load, or loads into a closed valve. The commonest causes on Indian sites:
| Failure mode | What happens | How the test exposes it |
|---|---|---|
| Standby not in the sequencer or remote-start loop | Duty trips; nothing tells the standby to start | Trip the duty machine with the standby left in auto, exactly as in service |
| Standby left in local or manual mode after maintenance | It waits for an operator who is not there | Check the mode selector before the test and record it |
| Auto-restart after power failure disabled | After a grid dip or DG changeover, neither machine restarts | Include a supply interruption in the annual test |
| Discharge or header isolation valve closed | Standby runs, builds pressure on its own side and unloads | Walk the line and record each valve position on the sheet |
| Dryer bypassed or not interlocked to the standby | Wet air reaches the process during changeover | Log dew point downstream during the test |
| Flat batteries or charger failure (diesel standby) | Engine cranks slowly or not at all | Check battery voltage at rest and during cranking |
| Seized or sticking inlet valve, minimum pressure valve or motor | Machine starts but cannot load, or trips on overload | Time from start to loaded; record motor current |
| Oil degraded or water-laden after long idle periods | High temperature trip; airend bearing damage | Oil condition check; discharge temperature trend |
| Condensation inside the idle machine | Corrosion, rust in the airend, wet separator element | Drain checks; inspect before the run |
| Capacity shortfall in summer | Standby runs fully loaded, but pressure keeps falling | Test in April or May at peak load and peak ambient |
None of these shows on a walk-round. All of them show in a loaded test.
A second compressor is only capacity on paper until a loaded changeover has shown that pressure at the critical user stays above its limit, and that evidence has to be renewed, because readiness decays every week a machine stands idle.
N+1 or partial redundancy: deciding from criticality
Not every plant needs a full-size standby. The right level depends on what happens when pressure is lost. Classify each air user, then size the standby to protect the users that matter.
| Criticality of the user | Consequence of losing air | Typical standby arrangement |
|---|---|---|
| Safety or process-critical (instrument air for valves, sterile or batch processes, furnace controls) | Unsafe state, scrapped batch, long restart | N+1 with automatic changeover, plus storage sized to ride through, or a dedicated priority supply |
| Production-critical (packing lines, presses, CNC clamping) | Lost output for the duration, restart time | N+1 or a standby covering the critical share, with priority valves shedding general use |
| Important but tolerant (general workshop, cleaning, tools) | Inconvenience; work can pause | Partial redundancy, or rental air on call |
| Non-critical (blowing, cleaning) | None that matters | Shed first; no standby needed |
In N+1, the plant can lose any one compressor and still meet peak demand at the required pressure. With three 10 m³/min (353.1 cfm) machines on a 17 m³/min (600.3 cfm) peak, the loss of one still leaves 20 m³/min (706.3 cfm), so the plant is N+1. With two 10 m³/min machines on the same 17 m³/min peak, it is not. Losing one leaves 10 m³/min against 17. That works only if 7 m³/min (247.2 cfm) of non-critical demand can be shed automatically. Write the classification down. Our criticality and service-level worksheet and the criticality classification guide set out the method.
When you size the standby, remember that it must carry peak demand at the summer inlet temperature, with its dryer and filters in circuit. A worn standby may also deliver 10–20 % below its nameplate. Verify its actual output with a pump-up test and capacity check.
The loaded changeover test, step by step
The test reproduces a real trip under real demand and measures what the critical user experiences. Run it during a normal production shift, preferably one of the busiest.
- Plan and permit. Agree the time with production. Name the critical user and its minimum acceptable pressure. Brief the operators. Have a person stationed at the duty machine ready to restart it if the standby fails. Follow the site’s isolation and permit-to-work rules for any work on the machine.
- Pre-checks on the standby. Record the mode selector (auto/remote) and the oil level and condition. Drain condensate. Check that the inlet filter, cooler fins and cabinet filters are clean. For a diesel standby, check fuel level, coolant level and battery voltage. Record running hours.
- Walk the line. Record the position of every valve from the standby discharge to the header: discharge, dryer inlet and outlet, dryer bypass, filter bypasses and header isolation. Confirm that the dryer the standby feeds is energised.
- Set up instruments. Fit a pressure logger at the critical user and one at the compressor house header, both at 1-second intervals. Fit a dew point sensor downstream of the dryer. Have a stopwatch or synchronised clock to time events, and a power or current reading on the standby.
- Record the starting state. Record the pressure at both loggers, the duty machine’s loaded or unloaded state, plant demand if you have a flow meter, and ambient and compressor room temperatures.
- Trip the duty machine. Use the method that represents a real fault: the emergency stop, a simulated fault, or switching off its supply. A normal stop may give the sequencer a warning that a real trip would not. Note the time to the second.
- Time the response. Record the time from trip to standby start signal, to motor running, and to loaded. For a star-delta machine, include the star-delta transition. For a diesel standby, include cranking and warm-up.
- Watch the critical user. Record the minimum pressure reached at the critical user and the time at which it occurred. Note whether any process alarm, machine stop or quality issue arose.
- Measure recovery. Record the time from the standby loading to pressure at the critical user returning to its normal minimum and then to the normal operating band.
- Run for stability. Keep the standby carrying the plant for at least 60 minutes. Record discharge temperature, oil temperature, motor current, any alarms, dew point and pressure stability. A machine that passes the first minute but trips on high temperature after 40 minutes has failed.
- Restore and debrief. Return the duty machine to service and confirm the system is back to normal. Record pass or fail, defects, and an owner and date for each action.
Pass criteria should be written before the test, not after. A typical set: the critical user’s pressure never falls below its stated minimum; the standby is loaded within the time the storage allows; pressure recovers to the normal band within an agreed time, often 2–3 minutes; no trip or alarm occurs in the 60-minute run; and the dew point stays within the dryer’s specification, for a refrigerated dryer +3 to +7 °C pressure dew point.
Will your receiver carry the plant through the changeover?
From the moment the duty machine trips until the standby delivers air, the plant runs on stored air. The calculator below answers whether that stored air lasts long enough. It uses a simple isothermal estimate:
- Usable free air (m³) = (V ÷ 1000) × (Ptrip − Plow) ÷ 1.013. V is the total receiver and pipe volume in litres, Ptrip the pressure when the duty machine trips and Plow the lowest acceptable pressure at the critical user, both in bar g. The figure 1.013 converts to free air at atmospheric pressure.
- Net withdrawal (m³/min) = plant demand Qd minus any supply Qs still available during changeover, for example from a second running machine.
- Ride-through (seconds) = usable free air ÷ net withdrawal × 60.
- Margin (seconds) = ride-through minus the standby’s time t from trip to delivering air. A negative margin means pressure at the critical user will fall below its limit.
- Volume needed for t (litres) = net × (t ÷ 60) × 1.013 ÷ (Ptrip − Plow) × 1000.
Take Ptrip as the lowest pressure in the duty machine’s normal band, because a trip can happen at any point in the cycle and the worst case sets the answer. Take Plow at the critical user, allowing for the pipe and filter drop between the receiver and that user. The estimate ignores the cooling of expanding air and the slow start of a standby’s delivery, so add a margin of at least 25 % to the result.
Worked example (illustrative)
Situation: total receiver and pipe volume V = 2,000 litres. The duty machine trips at its lower setting, Ptrip = 7.0 bar(g) (101.5 psig). The critical user needs at least Plow = 6.0 bar(g) (87.0 psig). Plant demand Qd = 8 m³/min (282.5 cfm). No other supply runs during changeover (Qs = 0). The standby takes t = 30 seconds from trip to delivering air.
Step 1: Usable free air. (2,000 ÷ 1,000) × (7.0 − 6.0) ÷ 1.013 = 2 × 1 ÷ 1.013 = 1.974 m³.
Step 2: Net withdrawal. 8 − 0 = 8 m³/min.
Step 3: Ride-through. 1.974 ÷ 8 × 60 = 14.8 seconds.
Step 4: Margin. 14.8 − 30 = −15.2 seconds. Pressure at the critical user falls below 6.0 bar(g) about 15 seconds before the standby delivers air.
Step 5: Volume needed for 30 seconds. 8 × (30 ÷ 60) × 1.013 ÷ 1.0 × 1,000 = 8 × 0.5 × 1.013 × 1,000 = 4,052 litres. The plant has 2,000 litres, which is about half.
Fix options, each recalculated with the same model:
- Larger receiver. Adding 3,000 litres (V = 5,000) gives 4.936 m³ usable and 37.0 s ride-through, a margin of +7.0 s. This is the simplest fix where there is space.
- Higher pre-trip band. Raising the duty machine’s lower setting to Ptrip = 7.5 bar(g) (108.8 psig) gives 2.962 m³ and 22.2 s, still 7.8 s short. On its own it needs 2,701 litres. Each 0.5 bar (7.3 psig) raise also costs about 3–3.5 % more energy every hour, so this should not be the only measure.
- Faster standby. If the standby is already running unloaded or is a VSD in hot standby and can deliver within t = 10 seconds, the margin becomes +4.8 s on the existing 2,000 litres. Only 1,351 litres is needed. The cost is unloaded running energy on the standby, so reserve this for critical processes.
- Critical-load priority valve. A pressure-sustaining valve that sheds 3 m³/min (105.9 cfm) of non-critical demand when header pressure falls reduces net withdrawal to 5 m³/min (176.6 cfm). Ride-through becomes 23.7 s, still short by 6.3 s. Combined with Ptrip = 7.5 bar(g), it gives 35.5 s and a margin of +5.5 s. For the combined case, needed storage drops to 1,688 litres, which the existing 2,000 litres covers. Shedding alone, at the original 7.0 bar(g), still needs 2,533 litres.
Confirm the chosen fix with a loaded test, because a real standby’s delivery ramps up rather than switching on instantly.
See also the receiver transient storage calculator.
How often to test, and how to rotate duty
| Check | Interval (suggested starting point) | Notes |
|---|---|---|
| Standby mode, valve positions, oil level, drain | Weekly | Two-minute walk-round with a signed checklist |
| Standby run on load for 30–60 minutes | Fortnightly to monthly | Brings oil and airend to temperature and drives off moisture |
| Full loaded changeover test with logging | Quarterly for critical plants; six-monthly otherwise | Use this page’s procedure and record sheet |
| Changeover with a supply interruption | Annually | Tests auto-restart after power failure and DG changeover |
| Peak-summer test | Annually in April or May | Proves capacity at the worst ambient |
| Duty/standby rotation | Every 500–1,000 running hours, or monthly | Keeps running hours balanced |
Aim to keep the running hours of the two machines within a few hundred of each other. Plan rotation so that both machines do not reach a major overhaul at the same time. If a sequencer handles rotation automatically, check every quarter that it really does. Record running and loaded hours in a compressor log book.
Preserving an idle machine
An idle standby needs active preservation: monsoon humidity condenses in a cooling airend, dust settles in cooler fins, and rodents find panels.
- Run it on load for at least 30–60 minutes, long enough to reach normal oil temperature, at least every two to four weeks.
- Drain condensate from the receiver, dryer and separator legs before and after each run.
- Check oil condition at each service and send a sample for analysis if the machine has stood through a monsoon. Milky or water-laden oil should be changed before the machine is relied on.
- Keep panel heaters powered where fitted, keep panel doors closed, and seal cable entries against rodents.
- For a diesel standby, keep batteries on a working float charger and check the electrolyte or voltage regularly. Turn fuel over, drain water from the fuel filter, and check coolant and engine oil.
- Service by the calendar as well as by hours. Oil, filters and separators age even when the hour meter barely moves. Follow the maker’s time-based intervals. See condition-based maintenance.
Rental air as standby
Where a permanent standby cannot be justified, or the existing standby is out for overhaul, rental air can provide cover. To make it work in a breakdown, prepare in advance:
- Fit a tie-in point on the header, with an isolation valve and a flange or coupling sized for the rental flow, outside the compressor house where a trailer can reach it. See tie-in and changeover.
- Agree the electrical supply, either a suitably rated outlet or a diesel unit that needs none. See electrical supply.
- Route rental air through a dryer and filters suited to your process.
- Keep the plant’s air requirement (flow, pressure, quality and hours) on file so that a request can be made in minutes. See breakdown rental and redundancy and failure management.
Rental air takes hours, not seconds, to arrive and connect: it is a planned second line of cover, never a substitute for storage or an automatic standby on a critical process.
Standby changeover test record
| Item | Value / observation | Pass criterion / note |
|---|---|---|
| Site, date, time, shift | ||
| Duty machine (make, model, serial, hours) | ||
| Standby machine (make, model, serial, hours) | ||
| Critical user and minimum pressure, bar g (psig) | Agreed before the test | |
| Standby mode selector position | Auto / remote | |
| Valve positions checked (list) | All open on the standby path; bypasses closed | |
| Oil level and condition; condensate drained | ||
| Battery voltage at rest / cranking (diesel) | ||
| Ambient and room temperature, °C | ||
| Header and critical-user pressure before trip, bar g | ||
| Plant demand at trip, m³/min (if metered) | ||
| Trip method and time | Represents a real fault | |
| Time to standby start / running / loaded, s | Within ride-through | |
| Minimum pressure at critical user, bar g, and time | Not below the agreed minimum | |
| Recovery time to normal band, s | Within the agreed time | |
| Alarms and process events | None | |
| Discharge / oil temperature at 15, 30, 60 min, °C | Within maker’s limits | |
| Motor current at 60 min, A | ||
| Pressure dew point during run, °C | Within dryer specification | |
| Result (pass / fail) and defects found | ||
| Corrective actions, owner, target date | ||
| Tested by / witnessed by / signatures |
Exceptions and cautions
- Do not test the changeover where a pressure dip could create a safety hazard, for example instrument air holding safety valves, until a back-up supply or an agreed safe state is in place.
- Agree the test with production and quality; some batch processes cannot tolerate even a controlled dip.
- Tripping a machine by its supply should be done only by an authorised electrician and within the site’s electrical safety rules.
- Pressure vessels at factories are covered by factory safety law. The Occupational Safety, Health and Working Conditions Code, 2020 took effect on 21 November 2025, and existing rules continue to apply during the transition. Before adding or relocating a receiver, confirm the current examination and testing requirements with your competent person or factory inspector.
Common mistakes
- Testing with a no-load start and calling the standby ready.
- Measuring pressure at the compressor house and not at the critical user.
- Testing in December and assuming the result holds in May.
- Leaving the standby in manual after maintenance.
- Not including the standby in the sequencer or remote-start circuit.
- Bypassing the dryer “temporarily” and forgetting it.
- Letting one machine reach 30,000 hours while the standby stays at 3,000, so the standby’s faults are never found.
- Relying on rental air for a process that cannot tolerate even a minute’s dip.
How Reddy Compressors helps
Reddy Compressors has been an independent multi-brand compressed-air specialist since 2002. We are suppliers, not dealers, and not an authorised dealer of any make. We have tie-ups with all major and reputed brands. Our service covers AMC, overhaul and spare parts for Atlas Copco, Ingersoll Rand, ELGi, KAESER, Chicago Pneumatic and other major makes, which covers the standby machine as well as the duty one. We supply new, used and refurbished compressors, receivers, dryers, filters and piping when a test shows a capacity or storage gap. Our own rental fleet, 50–2,000 cfm (1.4–56.6 m³/min) at 2–30 bar, provides temporary or breakdown cover. More than 20 service engineers and technicians work across Telangana, Andhra Pradesh and districts within about 300 km of Hyderabad, and we supply across India.
What to send us
- Make, model, FAD, pressure and running hours of the duty and standby machines
- Receiver sizes and an approximate main pipe length and diameter
- The critical user, its minimum pressure and the consequence of losing air
- Your last changeover test result, if any, and any recent trips
- Controller or sequencer details, and whether the standby runs on mains, DG or diesel
Use the service and maintenance enquiry. Call/WhatsApp 99590 60718 or 92468 88812.
Frequently asked questions
Is a no-load start of the standby compressor enough to prove it is ready?
No. A no-load start shows only that the motor turns. A readiness test must trip the duty machine under real demand and show that the standby starts, loads and holds pressure at the critical user, then runs stably for at least an hour.
Watch
Short explainers on this topic
Slide-by-slide engineering explainers from Reddy Compressors. More in the video library.
How often should a standby compressor changeover be tested?
A sensible starting point is a full loaded changeover test quarterly for critical plants and six-monthly for others, an annual test that includes a supply interruption, and an annual test in April or May at peak summer load. Weekly checks and a monthly loaded run fill the gaps.
What should be recorded during a changeover test?
Record the minimum pressure at the critical user and when it occurred, the time from trip to standby start and to loaded, recovery time to the normal band, any alarms, discharge and oil temperatures over a 60 minute run, motor current and pressure dew point downstream of the dryer.
Do I need a full N+1 standby compressor?
It depends on criticality. N+1 means the plant can lose any one compressor and still meet peak demand at the required pressure. Where only part of the demand is critical, a smaller standby with automatic shedding of non-critical users can be enough. Classify your users first.
How do I know if my receiver will carry the plant through the changeover?
Usable free air is about the volume in litres divided by 1,000, multiplied by the usable pressure drop in bar, divided by 1.013. Divide that by the net demand in cubic metres per minute and multiply by 60 to get seconds of ride-through. Compare it with the standby’s time to deliver air.
Why does a standby that worked in winter fail in summer?
Compressor rooms in Telangana and Andhra Pradesh can reach 45 to 50 degrees C. Hot inlet air reduces delivered air, plant demand is often higher, and an idle machine with dirty coolers or degraded oil can trip on high temperature. Test at peak summer conditions.
How should duty and standby compressors be rotated?
Swap duty and standby roughly every 500 to 1,000 running hours or monthly, keep their running hours within a few hundred of each other, and plan rotation so both machines do not reach a major overhaul at the same time. Check quarterly that any automatic rotation in the sequencer is actually working.
Can rental compressors be used as standby?
Yes, as a planned second line of cover, provided a tie-in point, electrical supply and air treatment are ready in advance. Rental air arrives in hours rather than seconds, so it cannot replace storage or an automatic standby for a process that cannot tolerate a pressure dip.
Related pages
Technical content reviewed by D. Srinivas Reddy, Managing Partner and service engineer, Reddy Compressors. Last updated 7 November 2026. How our engineering content is reviewed.