CONTRIBUTING RE.ENGINEER
Sanya Mathura of Strategic Reliability Solutions Ltd
CHALLENGE
A client wanted to extend the life of the oil in service by a further 6 months past the routine change date. This would allow them to complete their required level of productions to fulfill their customer’s orders. The extension of the oil life is carded to take place on their critical pieces of equipment for Train B (a series of 6 GE Frame 5 Gas turbines).
More background: The facility produces liquefied natural gas and has 4 trains each consisting of 6 or 7 GE Frame 5 turbines. Each train was commissioned separately and are treated as separate units with differing dates for end of service life. After seven years, each train is recommissioned with new turbines and a full overhaul completed on the entire train. For Train B, the time was approaching for the mandatory overhaul however, the company needed the extra capacity to fulfill customer orders. As such, it decided to extend the service life of the equipment on the train but needed to ensure that the lubricant could withstand the additional time and perform successfully.
SOLUTION
Baseline samples were taken from each of the turbines (critical units) on the train to establish the current condition of the lubricant in service. The following tests were performed; MPC, RPVOT, RULER, Viscosity, TAN, Tests for contaminants (including ISO 4406), additive packages and wear metals. It was found that the wear metals were considerably low which indicated that there
was minimal wear and that the lubricant was still protecting the elements. Viscosity and TAN levels
were well within the acceptable limits as well as the additive packages. The contaminants were
minimal and the MPC levels were below 15. However, the RPVOT and RULER tests indicated that the oil was nearing the end of it life showing sharp decreases within the past year from an average of 72% to 47% for most of the units.
With turbine oils, the warning flags are raised when the RPVOT or RULER levels get to 25% of the
original value. From trending the RPVOT and RULER values in turbine oils, it has been noticed that
both the RPVOT and RULER values are initially high then nose dive towards the 25-35% region and
taper off to remain in this region sometimes for years. As such, this drastic decrease was seen as
normal for the length of time that the oil had been in service (approximately 6 years). However,
given the risk involved of extending the life and the assurance needed for no unplanned shutdowns,
the team wanted to ensure higher RPVOT & RULER values.
To increase the RPVOT and RULER values, the oil could be sweetened however, the ratio of the new
oil would have to be determined. For the sweetening process, one must ensure that the new oil is
very similar to the oil in service. As such, one should verify if there were any formulation changes
from the batch that was used to initially fill the equipment compared to the tentative new fill.
Additionally, it should also be noted that even though the oil may be sweetened, this does not
indicate that the RPVOT or RULER levels will return to that of new oil. Given that the plant could not shut down, the sweetening of the oil had to be performed while the plant was operational. In this case, it was decided to sweeten the oil by 30%. To perform this task while the equipment is still running would indicate that an average of 30% of the sump must be drained. Before this was done, the OEM was consulted to determine the minimum sump levels that can be used without causing harm to the equipment while in operation. Thankfully, for this equipment the minimum levels were greater than 40%, thus a sump capacity of 70% during the sweetening process would not harm the equipment. Once the oil had been removed, the new charge of oil was pumped into the sump. Ideally, the same pumping equipment should not be used as this will transfer any contaminants from the old oi into the new oil.
RESULTS
Once the new charge was filled, we allowed the equipment to normalize and continue working for at least 24 hours before a baseline sample was taken to determine the new RPVOT and RULER values. Both values rose by approximately 20% (the new values were in the range of 68-75%) for the units which were sweetened. These values were monitored monthly to ensure that any drastic changes could be found in time and not allow for an unplanned shutdown.
**********
Contact Info:
Website: strategicreliabilitysolutions.com