Subscribe to our

Pumps & Systems, June 2008

A common mistake when troubleshooting a motor failure is to focus attention only on the motor. The perception is if the motor is failing, the problem must be the motor, but this is the wrong approach.

When assessing a motor failure, one should take the entire system into account. The motor, only one component of the system, reacts to the system demand. Theoretically, if the motor operates within design parameters in a clean environment, it will run trouble-free for many years. Unfortunately, this is not the reality-dirty, dusty, hot conditions compounded by loss of system control, voltage unbalance and power surges that can impact the life cycle of a motor.

Plant life cycle service is the fastest growing market opportunity in the industrial market today. Motor system energy optimization is a subset of these services and has tremendous growth potential for the U.S. and global pump industry. These types of life cycle services will be a fast-growing segment of the pump industry in the future and within 20 years will likely constitute 20 percent or more of the total pump industry revenue stream. The pump and allied industries has generally been slow to recognize and adapt to these new market realities.

The following case study incorporates a total systems approach to troubleshooting, specifically:

• Six Sigma root cause failure analysis
• Pump testing/evaluation
• Motor testing/evaluation
• Operations/procedures
• Installation, maintenance and repair procedures

Note: In the interest of space, the case study in this article only contains results pertinent to the final resolution.

The pumping system in question (three pumps in parallel) was installed in the early 1970s. Subject units were between bearing, single stage, double suction pumps with side/side suction discharge. Pumps were directly coupled to a fixed speed, 2,000-hp, 1,780-rpm induction motor with a limited end float gear coupling. The fixed speed pumping system developed 850-ft of head at 8,000-gpm when operated at design point (full load) with two pumps in service (the third pump was installed spare).

Current Situation

Condensate booster pump motors were running at a higher temperature than the original units (replaced in the 1990s). The stator temperature trending indicated the motors continued to operate at a higher temperature than their initial installation. Inboard and outboard motor bearings on three Unit 1 Condensate Booster Pump Motors were failing prematurely.

If this was a motor problem, why was the stator temperature so high in comparison to load? Was there a system problem? Was this a motor-to-pump alignment issue? Did the customer receive the correct replacement motor? Was this a combination of issues?

Gathering the Facts

After reviewing historical and current DCS data and communicating with operations and maintenance personnel, we built an overall picture of the situation. From initial commissioning until the late 1980s, the plant ran at reduced load; therefore, subject pumps saw intermittent full load conditions. In the late 1980s, the plant operation changed, which forced the condensate booster system to operate closer to design point.

The general perception from plant personnel was that the original motors delivered reliable service. However, after further discussion with electrical maintenance personnel, we found the plant began experiencing cracked rotor bars in the original motors, which prompted their replacement in the early 1990s. The new motor vendor supplied six replacement motors during 2 years. Another operational change occurred about the same time as the motor replacement process, and the plant began operating at design point 24/7.

Within a few months of the first motor installation, the plant noticed the stator temperature was higher than that of the original motors. According to plant personnel, this elevated stator and bearing temperature existed on all six motors and became most apparent during the summer. When the new motor installations occurred, the stator temps were in the range of 250- to 280-deg F and became as high as 325-deg F. The high stator temperature had not been the direct result of a motor failure or subsequent multiple rewinds.

No operational or maintenance/design changes had been made in the last few years, with the exception of a valve change. 

The motors were across the line start. Normal operation was to have two of the three motors in service, with the third auto-starting if the pressure drops on the in-service pumps. Other than removal of a motor/pump for maintenance, the starts had been a monthly check of standby equipment. The "off" motor/pump had been rotated monthly. There was not a defined start-up procedure. Pumps were "valved" in allowing immediate flow when needed. Most of the pump problems had been typical bearings/seal problems, and the units had been repaired numerous times through the years. While the customer could trend stator and bearing temperatures, they could not track amperage or individual pump flows under normal operation.

The coupling was a size 4.5-in limited end float gear type. Pump and motor vibration data had been collected monthly by staff personnel. The plant did not maintain records of pump and motor repair, and such repairs were typically performed in-house by plant personnel. The plant did not track how many times a given motor had been rewound, nor did they performance-test equipment after repair.

Interviews with plant personnel provided us with the following information:

• Typical reason for motor repair was bearing (sleeve type) failure.
• Pumps had never been removed for major overhaul-only pulled rotating element for repair.
• Pump replacement parts had been supplied by both the OEM and aftermarket suppliers.
• Alignment practices (off sets) for the replacement "hot" motor were the same as the original motors.
• The customer's "sister" plant (Unit 3) had a similar system/layout. The sister plant still had the original motors in service, and the units ran reliably-with the exception of pump motor 3, which required yearly repair due to cracked rotor bars. In one unique difference, the sister plant operated all three pumps in parallel, which resulted in lower amperage draw and reduced stator temperatures.
• Sister plant also tested the five vane impellers and then chose to revert back to six vane (it is unclear exactly why).

Based on this information, we developed a "Fishbone" diagram to identify potential cause and effect. Typically, when performing a failure analysis, one would focus on a particular mode of failure, i.e., stator temperature or bearing failure. In this particular case, the team chose to capture both the bearing failures and stator temperature together.

Analysis/Verification Plan

Thermal imaging of the motors indicated hot spots at the outboard bearing and center of stator. Imaging readings were on the high side of acceptable operating range. The motor with the highest reading was Pump 1.

Each pump was tested individually and then in parallel to verify individual and system performance. Vibration data was captured at each test point.

Potential Root Cause Verification

• Motor Undersized For Pump-Testing confirmed, with two pumps in service, valves 100 percent open, each pump delivered 8,000-gpm at 850-ft head and required 2085-hp. The performance on Pump 3 indicated unit was working harder than the other two units.
• Phase Imbalance-Testing confirmed that phase imbalance was less than 1 percent.
• Load/System Operational Change-Historical data and interviews with plant personnel verified from initial commissioning until the late 1980s the plant ran at reduced load; therefore, subject pumps saw brief intermittent full load conditions. In the late 1980s, the plant operation changed, and this operational change forced the condensate booster system to operate at design point 24/7. Demand change pushed the motors into their service factor.
• Motor/Pump Alignment-Bearing damage indicated potential misalignment, and vibration data indicated the same. Misalignment can result in elevated bearing temperature, oil degradation and premature bearing failure.
• Quality of Motor Rebuilds-Motor 3 was drawing higher amperage than other units, which could be due to the quality of rewind. Additionally, rewinds can result in reduced motor efficiency and higher stator temperatures.
• Motors Operating in Service Factor-Induction motors are designed to operate within designated service factor; however, loading, ambient temperature and power factor will affect the thermal rise. Maximum temperature limit is 311-deg F, Motor 3 tended to operate in this range 80 percent of the time.
• Multiple Pump Operation-Test data indicated two pumps in parallel operation require 2,032-hp, forcing motor into its service factor. With three pumps in service, the horsepower was well within motor design range as well as pump acceptable operating range.