High current at start-up and a cone that would not clear the impeller: the parts were right, the pairing was not

At-a-Glance

The Setup

A process plant was commissioning identical cone-inlet centrifugal blowers on the same duty — 36,000 CMH at 60 mmWC, on 25 HP motors. Under the supply plan agreed for this order, the impeller assemblies and the casings travelled to site separately, to be married up there during erection.

That detail matters. A centrifugal fan is not a bag of interchangeable parts: each impeller is balanced inside its casing, with its inlet cone clearance set against its impeller shroud. On paper, sister machines of the same rating look identical. In the field they are not — each is a matched set, and the tolerances that keep a fan quiet live in the pairing. And these blowers sat on the critical path of a plant start-up.

The Complication

The site reported the machines drawing high current on start-up and refusing to come up normally. The plant team then spotted a placement error themselves, moved a blower to its correct position — and that unit's current dropped into the desired range.

But it did not end there. On that same machine, with the inlet cone installed, the site now reported vibration and noise, with the cone-to-impeller clearance visibly tight. And on the other machine, current came into tolerance only with the cone removed — the cone rubbed hard against the impeller and physically could not be fitted. The site's own words: "the cone is having friction with impeller hence not installed."

The obvious suspect was the fan itself: a mis-built casing, an out-of-round impeller, a wrong rotation mark, an unbalanced rotor. That is the natural conclusion when sister machines misbehave in different ways on the same day. It was also wrong.

The Diagnosis & Fix

We worked it in order, ruling out the cheapest-to-check causes first.

One — were the parts built right? We went back to the manufacturing and dispatch record and compared it against the approved GA. Every casing and every impeller assembly was built correct to drawing, and the nameplate rotation arrow was correct. Critically, each casing had been marked with its own job identity before dispatch — visible in the dispatch photographs taken at our works.

Two — was it electrical? High current invites a motor, starter or supply diagnosis. But the site's own intervention had already answered that: simply re-locating a machine dropped its current into range, with no electrical component touched. A current that responds to where a fan physically sits is not an electrical fault.

Three — what did the site photographs and video actually show? The answer. The separately-shipped casings and impeller assemblies had been cross-matched during site assembly — a casing married to the impeller assembly belonging to its sister machine. That one error explains both symptoms: the wrong pairing loads the motor abnormally, and it destroys the inlet-cone-to-impeller clearance set at the works — so on one machine the cone runs tight and noisy, and on the other it fouls outright and will not go on.

The fix followed the cause. The plant re-located the units correctly, restoring current to range on the machine they corrected. We then sent a service engineer with an on-site dynamic balancing machine, tooling, a spare bearing and a belt, to trim the residual vibration, check bearing and drive condition, and reset the cone clearance. We carried the cost of that visit ourselves, though our standard policy is to charge for visits arising from faulty commissioning.

The Result

Here is what the record does — and does not — show.

• What we proved: every casing, every impeller assembly and the rotation marking were built correct to the approved GA, and each casing carried its identity mark at dispatch — evidenced in our own dispatch photographs. The fault was not in the parts as built.
• What was confirmed: the site's re-location of a wrongly-placed machine brought its motor current back into the desired range — confirmed by the customer.
• What we changed: we deputed field service with a balancing machine, tooling and spares, at our cost, to close out the residual vibration and cone clearance on site.
• What the record does not confirm: there is no final customer sign-off on file closing out vibration and cone fitment after the balancing visit, and no field vibration reading against an acceptance limit was captured — so we claim neither. (ISO 14694 rigid-mount BV-3 limits are 4.5 accept / 7.1 alarm / 9.0 shutdown mm/s; no measured site value exists here to compare against.)
• What we fixed for good: the systemic lesson went into our failure-mode library. Where casing and impeller ship separately, they now travel as an unambiguously identified matched set, with a one-page site marry-up instruction covering rotation and cone-clearance checks and a do-not-run-until-verified sign-off.

The Takeaway + Call to Action

Sister machines, different symptoms, one cause — and the cause was in the marriage, not the metal. When a fan misbehaves at start-up, before it has run a single production hour, the odds weigh heavily toward how it was received, assembled and installed, not how it was built. The discipline that pays is boring and cheap: check the identity marking, check the rotation, check the clearance, then press start.

(Our balancing is to ISO 21940; performance is tested to IS 4894 / ISO 5801 / AMCA 210 method; vibration is judged against ISO 14694.)

Commissioning a fan of any make that pulls high current, rubs or vibrates? Call us in. You will get an ordered diagnosis and a written corrective action, not a parts quotation.

— Jitamitra Electro Engineering · Technical Services

Engineered for Every Application.