If best scenario had played out there would have been additional cost savings of $ 2.8 Million

This case study illustrates how PROGNOST®-Predictor detected impacting in the time waveform of the gear pump along with a ticking sound coming from the gearbox.

Sequence of Events

  • 3/22/2018 at 9:30:00 am: PROGNOST Systems received message from customer who was concerned that they heard a ticking coming from Gear Pump gearbox. Customer was asked to look at the time waveform on one of the sensors near the ticking and we found a very strong impacting happening that was very concerning.
  • 9:38:00 am – Customer sent a screenshot of time waveform which showed a large amount of impacting that he had found. This impacting didn’t show really any great amount of increase in energy in the trend and was concerning why PROGNOST®-Predictor had not alarmed
  • At 9:50:25 am – It was noted that the machine was in a fixed speed mode and that TSA gear impacting was not collecting data and would not collect data in fixed speed. For TSA to work it must have a tach pulse which was not working properly.
Gear Impacting in Time Waveform
Gear Impacting in Time Waveform – Output Speed should be approx. 20.5 rpm
Output Speed



Gear Mesh 1508.25 cpm = 1508.25/75t
20.11 rpm is the Output Speed
  • At 10:30:00 am – It was decided that after the weekend the customer would open an inspection port in this gearbox and see what they can see.
  • On 03/26/2018 Monday morning they opened the inspection port and could only see the face of the gearing and saw no damage on the teeth as it was rolled over. They couldn’t see the side of the gear only the teeth face.
  • They decided that they would remove the cover for repairs the following week as PROGNOST was very concerned about the impacting being seen.
  • That following weekend before they were to open the box for repairs there was a significate step change in PROGNOST®-Predictor with the unclassified and with handheld readings that was very concerning to the customer.
Step Change
This step change could have been when the gear crack split all the way to the gear teeth.
  • On April 3rd the following week the customer opened the gear pump to the best of their ability, but they couldn’t get the cover off as it was sealed like almost welded on after 22 years of service and had never been opened.
  • As they were inspecting the box they now could see the damage as the Gear 6 was showing a cracking from the shaft to the teeth and around the radial of the gear. There was also significant gear tooth damage.
  • Gear 6
    • 75 teeth
    • 44 inches in diameter
    • 13 inches in width
  • As the customer couldn’t get the cover off it was then decided to ship it out to Philly Gear in Houston for repairs later that week.

Machine Specifications

  • Kobe, JSW
  • Specified Rating 1206 hp
  • In service for 22 years without removing cover
  • Input shaft speed 115 – 1150 rpm
  • Output shaft speed 4.5 – 46 rpm
  • Reduction ratio 1 : 25
  • Weight 21600 lbs
  • Oil quantity (lower casing) approx. 96 gals

Sensor Locations on the Gear Pump

Sensor Locations
Gear Pump / Gear Reducer with different sensor positions

Immediate Actions Taken

  • Box was inspected from an inspection port on Monday the 26th of March and nothing was seen.
  • It was decided later that day that they would shutdown and remove the cover for repairs the following week.
  • On April 3rd the customer was unable to remove cover as it was almost welded/cemented on since it was 22 years since this box was open, but they were able to see enough and discovered the cracked gear and damaged gear teeth.
  • Since customer couldn’t get cover off the entire gearbox was shipped to Philly Gear in Houston later that week for repairs.
  • Since the Engineer was monitoring this very closely and took action the damage was limited to only that this cracked gear. No other damage to the box
Cracked Gear 1
Images of Cracked Gear

Contributing Causes and Corrective Actions

  • 22 years of operation without being opened
  • Just normal maintenance was being done, but never opened
  • Repair gears and re-assemble
  • It is noted that the gearing will be what is called “soft gears” as they do not want to take the additional time to harden them (2 weeks)
  • Shipped gearbox back to customer to be re-installed
  • Install, align, add sensors, and fix tach sensor that was never working properly.
  • Startup and monitor for any issues that might a rise

Kurtosis Simulation (offline analysis)

Kurtosis Simulation Offline Analysis
These two TSA plots are computed offline to simulate what Kurtosis would look like if the trigger had been working properly

Cost View and Scenario Comparison

There are three different scenarios which are possible

No PROGNOST®-Predictor system and/or handheld monitoring. 20 weeks downtime. Box destroyed and needs to be replaced. Calculated Costs $31 Million
PROGNOST®-Predictor and handheld see the issue after the ticking was discovered. 4 weeks downtime. Calculated Costs $5 Million – this Scenario is what played out
TSA tach adjusted and working properly. PROGNOST®-Predictor sees impacting back in December and Customer has several months to plan for repairs.
Calculated Costs $2.2 Million

If best scenario had played out there would have been additional cost savings of $ 2.8 Million

Review

  • Several different pieces of information were used to determine that there was a major problem and to ultimately shutting down in a timely fashion.
  • The Vibration Engineer routinely walks the equipment, he routinely monitors the PROGNOST®-Predictor system, he used the time waveform signatures, used the unclassified spectrum, and along with the handheld data collector data made a very well informed decision.
  • PROGNOST Systems Houston was called for help and a second opinion on the situation.
  • Without all of these working together the damaged could have been very significant at the cost of approximately 31 million.