Being a power quality detective involves more than just finding a cause for an incident. Often, the job entails politics and communication with outside vendors and suppliers.

Alan, a facilities engineer for a well-known metropolitan museum, asked me to investigate a recent over voltage tripping problem occurring with his variable-frequency drives (VFDs) that controlled the air-handling units for the air conditioning to the galleries. With millions of dollars of artwork on display AC is used for more than a comfort, it is an absolute necessity to control humidity. You can just imagine what those 200-year-old oil paintings would look like on a hot, humid Baltimore summer day. As soon as he told me the problem kept showing up on several VFD units, I knew he had a change in his power source.

In PQ troubleshooting, it is always important to ask yourself, "Is this a local or global type of problem?" Since multiple units throughout the facility were experiencing the same faults, this had to be a global problem. It was most likely linked to the power source. I had started up these drives more than eight years ago and knew the installation well. In fact, they had performed exceptionally well and were under contract by me for yearly preventive maintenance.

VFDs are sensitive to voltage fluctuations and typically fault on over- or under-voltage conditions greater than 10% of nominal voltage. Over the eight-year period, he only had a handful of faults with these drives - primarily due to weather conditions. I asked Alan to have the local utility representative check the museum's service and verify any changes. That's when the real problems started.

Utility personnel brought in monitoring equipment and connected it to three problematic units. After monitoring the equipment for a week, the utility staff decided the drives were oversensitive. They recommended that the museum spend several thousand dollars to install isolation transformers to correct the problem. Of course, Alan and I were somewhat skeptical of this expensive solution. The units had performed exceptionally well for eight years, so why make a change now?

Next, we pulled out all of the yearly power-quality surveys for the VFDs and looked at the incoming line voltages over the last five years. On an average for all units, the nominal incoming voltage was 467V across all three phases. The present voltage at the units had risen to 490V during peak operating hours and well above 500V during non-peak hours. At this point, the culprit was anybody's guess. But we suspected the utility had something to do with this problem.

After gathering the facts, we asked to meet with a senior representative of the utility to discuss our concerns. As we soon found out, this area of the city continued to experience extreme problems with demand. The utility had set up a new capacitance switching arrangement to offset the problem. The net result was elevated voltages on the feeders to the museum, which of course affected the drives - thus causing unnecessary trips.

In this case, the history of the units over an eight-year period had a major impact on the outcome of the meeting. With the correct information and detailed records, we proved the drives had performed successfully for eight years, and that voltages to the units had changed. The utility had no choice but to back down and correct the problem. There is no doubt about it: The detailed power-quality reports maintained by the museum were a major factor in the utility's decision to take responsibility.

However, not all facilities are so lucky. Many times, the long-term benefits of implementing and maintaining power-quality studies for a building or facility are not recognized. Typically, this data is only thought of when problems begin to appear on the system and troubleshooting work begins. It's kind of like yearly checkups at the doctor's office. Are they really necessary? You bet. It's the same with your building.

Changes take place all the time in any facility, as well as at your neighbor's house across the street. This is where a well-documented yearly survey can make all the difference in coming to terms with an equipment manufacturer, contractor, or utility. At minimum, a building survey should include the information noted in the Table, on page 16.

While there are many other readings you can take, these are some of the most important when evaluating equipment-related power problems. Remember, measurements are most useful when taken during peak and non-peak periods. The differences between the two can be surprising. In fact, many buildings today have extremely excessive voltage levels during off-peak periods.

While this may not trip or shut down the equipment, these elevated voltages will shorten the life of many solid-state devices and power supplies. The difference between summer and winter months can also be revealing. If possible, you should consider two semiannual surveys to decrease your chances of problems.

The key to any good power-quality reporting system is consistency. Take the same measurements at the same time periods with the same loads in operation. Then, use the data to identify any significant trends or directions. In the end, any reporting is better than no reporting - even if you start off small and add to your reports as you can.

Sidebar: Case in Point

It is suggested that a large number of motor winding failures can be traced to a voltage imbalance of the three phases of supply to the motor. Even an imbalance as small as 2% to 3% will affect the torque capabilities of the motor, which can cause excessive heating on a fully loaded motor. So how can we protect ourselves from this imbalance? Here's a step-by-step procedure to check for voltage imbalance on any 3-phase piece of equipment, motor starter, VFD, UPS, etc.

Step 1: Measure voltage at each phase of the incoming power supply to the equipment in question. Readings are made from L1 to L2, L2 to L3, and L3 to L1.

Step 2: Add the voltages together.

Step 3: Divide the total by 3 to find the average value.

Step 4: Subtract the average voltage from the largest voltage reading to find the voltage deviation.

Step 5: To find the percent of imbalance, apply the following formula:

Voltage Imbalance = Voltage Deviation/Average Voltagex2 100

Example: Step 1: L1 to L24442V L2 to L34474V L3 to L14456V

Step 2: 442V`474V`456V 41372V

Step 3: Average voltage 41372V 3 34457V

Step 4: Voltage deviation 4474V (highest) 1 457V (average) 417V

Step 5: Voltage imbalance 417V 3 457V 2 10043.72%. This is great enough to severely affect the torque producing capability of a 3-phase motor - causing it to overload at full load conditions.

If you find an imbalance of 2% or more, you should consider the following:

A. Check for excessive loads to a single phase of the 3-phase source.

B. Consider reducing the load to the motor, if the imbalance cause cannot be identified.

C. Consult with your utility if the problem appears to be coming from the utility source.