Why failed interventions deserve as much attention as successes
Every experienced compressor engineer has seen a bigger machine installed that did not fix low pressure, a variable-speed drive that saved far less than promised, or a new dryer that still delivered wet air. These outcomes are expensive, but they are also the best teachers. They show which assumptions were wrong and what should have been measured first.
Suppliers rarely publish them. We think plant engineers benefit more from understanding why common fixes fail than from another set of success stories, so this page sets out how to analyse an unsuccessful intervention and describes the patterns we meet most often in our service and audit work. The examples are typical engineering patterns, not records of particular customers.
Five ways an intervention can fall short
| Status | What it means | Typical example |
|---|---|---|
| Objective not achieved | The target was missed entirely | Critical-user pressure still drops at peak after a compressor upgrade |
| Partial benefit | Some improvement, below the agreed threshold | Leak repair reduces unloaded running but not enough to retire a machine |
| Adverse side effect | Target met but something else got worse | Higher set pressure fixes one user but raises power and leak flow |
| Unsafe condition | The change created a risk | A bypassed dryer or filter sends wet, oily air to a sensitive process |
| Inconclusive | Evidence is too weak to decide | No baseline was logged, so a lower bill cannot be attributed |
An intervention can achieve its engineering goal yet fail a commercial payback expectation. State which criterion failed.
Start from the original hypothesis
A useful review begins by writing down what was expected before the work: the problem, the proposed mechanism, the expected change, the acceptance threshold and the evidence plan. If none of this was recorded, reconstruct it from quotations, meeting notes and emails, and say that it was reconstructed. A pipe replacement intended to reduce header pressure drop must be judged on pressure drop, not on how neat the new pipework looks.
Separate observation, hypothesis and finding
- Observation: “Pressure at the moulding shop fell below 5.5 bar (79.8 psig) during the second shift.”
- Hypothesis: “The branch to the moulding shop is undersized.”
- Finding: “Simultaneous logging at the station and the shop showed the loss occurred across a partly closed isolation valve.”
Only the finding can be treated as the cause, and only when the supporting tests are adequate. Many failed fixes started by acting on a hypothesis as if it were a finding.
A diagnostic sequence for a fix that did not work
- Confirm the change was made as specified: correct set points, valves open, control wiring and parameters as intended.
- Check the measurements: instruments, positions, time alignment and whether the baseline and after periods are comparable.
- Re-examine demand: has production, shift pattern or equipment changed since the baseline?
- Map pressure from the compressor discharge through dryer, filters, receiver and header to the critical user, logged at the same time.
- Check air treatment: dryer loading, inlet temperature, drain operation and filter differential pressure.
- List rival explanations and test each one before settling on a cause.
- Record the corrective decision and what will be measured to confirm it.
Common patterns we see in the field
Raising pressure that does not fix the end user
If a critical user is short of pressure because of a restriction in the distribution, raising compressor set pressure only pushes more air through the restriction while increasing power, typically about 7% for each extra bar, and increasing leak flow everywhere. The fix is to find and remove the restriction, or to supply that user locally.
A bigger compressor without a demand study
A larger machine bought because “pressure is low” often runs unloaded much of the time, because the real problem was peak demand, receiver capacity, pressure drop or leakage. Logging flow and pressure first usually shows whether more capacity is needed at all.
A variable-speed drive with disappointing savings
A VSD saves energy when demand varies within its efficient speed range. If the plant runs at a steady high load, or the VSD machine is not the trim machine in the sequence, savings can be small. A controller that lets fixed-speed machines run unloaded alongside the VSD can erase the benefit.
Leak repair with little measured benefit
If compressors do not unload or stop when demand falls, a lower leak rate shows up as more unloaded running rather than less energy. Leaks also return quickly without a repair register and re-survey. The repair worked; the system did not convert it into a saving.
A new dryer that still delivers wet air
Persistent water after a dryer upgrade usually points to high inlet temperature, an undersized or overloaded dryer at summer conditions, failed drains, bypass valves left open, or water entering downstream from cold pipework or unprotected drops. A refrigerated dryer typically gives a pressure dew point of about +3 °C only within its rated inlet and ambient conditions; in a hot Hyderabad compressor house these limits are often exceeded.
Additional receiver without stable pressure
A receiver helps only if the pressure band across it allows the stored air to be used and the connections are not restrictive. A large receiver behind a narrow connection or a slow control band does little for fast peak events.
Writing the lesson down
A lesson is only useful if the next engineer can find and apply it. We record each unsuccessful intervention in a short, consistent format:
- The original problem and the hypothesis that led to the change
- The acceptance criterion and how it was to be measured
- What was actually done, with dates and settings
- What was measured afterwards, and how it compared with the baseline
- The confirmed cause of the shortfall and any contributing conditions
- The corrective decision and the evidence that confirmed it worked
- What should be checked first on a similar job in future
Where a customer agrees to share a lesson publicly, it is published anonymously, without client names, prices or details that could identify the site, and only as far as the records support.
Root cause versus contributing condition
Most failures have one main cause and several contributing conditions. A dryer may be the right size on paper, but a blocked aftercooler, a summer ambient of 45 °C and a stuck drain together overload it. Fixing only one may not be enough. Record each condition, its evidence and whether it was corrected, so the next intervention starts from what is known rather than what was assumed. A short list of confirmed conditions is worth more than a long report built on guesses.
How Reddy Compressors helps
When a previous fix has not worked, we start by measuring rather than replacing. Our engineers log pressure at the station and critical users, flow, power and, where relevant, dew point, then trace the system from compressor to point of use. As an independent multi-brand specialist working on ELGi, Atlas Copco, Ingersoll Rand, KAESER, Chicago Pneumatic and other makes, we have no reason to push a particular brand’s machine as the answer.
If measurement shows that more capacity is needed, we can sell new or used/refurbished compressors, or supply rental air from our own fleet while the permanent solution is decided. Our service teams cover the whole of Telangana and Andhra Pradesh from our Tarnaka, Secunderabad office, and we supply equipment and spares across India. Call +91 99590 60718 or +91 92468 88812.
What to send us
- What was changed, when, and what it was expected to achieve
- The symptom that remains: low pressure, wet air, high energy, tripping
- Compressor makes, models and ratings, dryers and receivers on site
- Any logs, bills or controller data from before and after the change
- A simple sketch of the compressor house and main distribution
Frequently asked questions
Why is my air pressure still low after buying a bigger compressor?
Low pressure at a user is often caused by pressure drop in dryers, filters, hoses or undersized pipework, by short peak demands, or by leaks, not by lack of compressor capacity. A bigger machine cannot push air through a restriction any better. Log pressure at the compressor and at the user simultaneously to find where the pressure is being lost.
Why did my VSD compressor not save as much energy as expected?
Savings depend on how much demand varies and how the VSD machine is sequenced with others. At steady full load a VSD offers little advantage. If fixed-speed machines keep running unloaded beside it, or it is not the trim machine, savings shrink further. A demand profile logged before purchase shows whether a VSD suits your plant.
Why is there still water in my air lines after installing a new dryer?
Common causes are inlet air hotter than the dryer’s rating, an undersized dryer for summer conditions, faulty or blocked drains, a bypass valve left open, or condensation in cold pipework downstream. Check dryer inlet temperature, pressure dew point at the outlet and drain operation first. Water can also enter through a parallel undried supply such as a rented compressor.
We repaired our leaks, so why has the electricity bill not fallen?
If the compressors do not unload or switch off as demand falls, the saved air simply becomes more unloaded running, which still uses substantial power. Controls and sequencing need to match the lower demand. Leaks also return within months without regular surveys and a repair register, so the benefit can fade unless it is maintained.
How do I know whether an intervention actually failed?
Compare the outcome with the acceptance criterion that was agreed before the work, using measurements at the same boundary and under comparable conditions. If there was no criterion or no baseline, the honest conclusion is often “inconclusive” rather than “failed”. Define the target and the measurement plan before the next change, so the result can be judged.
Should I raise compressor pressure to fix one machine on the line?
Usually not. Raising generation pressure increases power, typically about 7% for each extra bar, and increases leak flow across the whole plant. If one user needs more pressure, check its supply pipework, hose and connections first, or consider a local booster or dedicated supply. Fix the restriction rather than penalising the entire system.
Related pages
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.