Using Risk and Condition Assessment to Drive Motor Repair Decisions

The Central Arizona Project (CAP) operates and maintains 109 motor/pump units in 15 pumping plants scattered across 336 miles of aqueduct. These units are indispensable to CAP’s mission of delivering Colorado River water to the people and farms of central Arizona.

CAP motors vary in size from 700 hp to 60,000 hp. Most were installed in the early- to mid-1980s and 25 percent of them have not been overhauled for at least 20 years. Given their criticality and age, CAP maintains an active motor condition monitoring program, including power factor tests, online partial discharge measurements and motor circuit evaluations.

One of CAP’s motors, a 5,200 hp Salt-Gila Unit 5 (SGL 5) synchronous motor manufactured in 1984, had some very high power factor results and increasing partial discharge activity. Because these tests are indicative of stator winding insulation deterioration, CAP scheduled a service outage for SGL 5 to evaluate and address the situation.

Figure 1: SGL 5 stator after reconditioning

In December 2015, the SGL 5 was disassembled and shipped to an outside motor shop to perform a complete electrical evaluation. Just before Christmas, the shop recommended that CAP rewind the SGL 5 stator at a cost of around $225,000.

This posed a problem for CAP management because its board of directors had just approved the 2016 budget, which did not include funds for a rewind. Based on past practices after similar test results, the CAP maintenance staff had budgeted only $30,000 for SGL 5, the projected cost of reconditioning the stator.

So, management was faced with a tough decision: go back to the board of directors and ask for more money before the new budget period had even begun, or proceed with reconditioning as budgeted even though the outside contractor recommended rewinding? To assist in that decision, management posed these questions to its operations and maintenance personnel:

• If CAP reconditioned SGL 5, rather than rewinding, and the unit were to fail as CAP entered its peak summer pumping season, would we still be able to make all scheduled customer deliveries?
• How do the latest test readings on SGL 5 compare to past readings on that unit and to other similar units across the CAP system?

Operations responded that CAP could still meet its scheduled deliveries even if it were to lose SGL 5 in the summer of 2016. That gave management some flexibility in deciding what to do with the unit.

Interestingly, the second question posed more of a challenge. Although CAP had been collecting motor test data for at least a decade, there was no easy way to compare that data across units and time. But CAP’s reliability engineers responded by compiling all motor test results into a central database and developing a system for ranking the condition of all 109 motors. The data and ranking system showed that SGL 5 was in better condition than was previously thought. Although some power factor measurements were very poor, they did not necessarily indicate the need to rewind the stator. The partial discharge measurements were also less concerning, as the ranking system indicated the condition was only moderately deteriorated. Also, the partial discharge test results were significantly higher for all 10 of CAP’s motors compared to its other motors, leading management to conclude that the ranking method may need to be adjusted for that group.

With this information in hand, management made the decision to proceed with reconditioning SGL 5 as originally planned and budgeted, but directed the maintenance staff to test and monitor the unit after the repairs were completed to evaluate their effectiveness. In this way, CAP would use SGL 5 as a test bed to inform future decisions for the rest of its aging motor fleet. If the reconditioning was successful, CAP might be able to prolong the useful life of other motors at a relatively low cost; if not, then CAP would need to begin budgeting for more costly rewinds in the near future.

So how did the reconditioning of SGL 5 turn out? The power factor readings showed dramatic improvement over 2015 levels after the unit was reconditioned. Partial discharge is being trended before reaching a definitive conclusion.

In Figures 3-5, the background color gradations approximate the rating criteria developed based on the engineering database, with green indicating “like new,” blue “aged,” yellow “degraded” and red “failing.” The “failing” ranking means that CAP believes it has reached the “P” point in the P-F interval.

Two different power factor measurements are performed on each stator winding, a grounded specimen test (GST) and an ungrounded specimen test (UST). The GST power factor assesses the insulation between the winding and ground (i.e., the core). This is the majority of the insulation in the motor and a failure would likely require major repair or a complete rewind. The UST power factor assesses the insulation between phases, which is primarily the end winding. The GST power factor is the best indicator of overall insulation condition. For SGL 5, the GST readings were good even before reconditioning. Following repair, the GST reading was about one percent, better than it had been in a decade.

The UST power factor trend for SGL 5 was considerably worse than the GST results and had been high for the last five years. These are the results that really drove the decision to recondition the motor.

The latest power factor measurements after reconditioning were better than they were 10 years ago. A few were actually slightly negative, which is not unusual and is caused by the semi-conductive paint used to minimize the voltage stress at the point where the coil exits the core slot.

Partial discharge is a measure of electrical discharges in the stator winding, indicating voids in the insulation. The manufacturer of the test equipment CAP uses, recommends using the Q_m value as the first criterion for evaluating the insulation condition. Q_m refers to the highest magnitude discharges that occur with a frequency of at least 10 per second. The blue/yellow boundary is where at least 75 percent of all similar voltage motors in the manufacturer’s database have Q_m readings below that value. The trend is slowly increasing and has a few flier points, but overall the condition is good.

The reason for measuring power factor and partial discharge on motors is to gain information on the condition of the insulation. When the measurements indicate problems are developing, the intent is to understand what the problems are and develop and execute an appropriate plan to correct those problems.

In this case, CAP had clear indications of a developing problem with SGL 5. Alternatives were evaluated, taking into account operational risk, and a plan was developed and executed. So far, measurements taken after the work was completed show that it was highly successful in mitigating the problem.

CAP is continuing to monitor the unit with semiannual testing to validate the results.

Jeff Guy

Jeff Guy has worked as an electrical engineer at Central Arizona Project (CAP) for over 30 years. Jeff has been a reliability engineer since the group was formed in 2013. Prior to that, he was a maintenance engineer.
www.cap-az.com