Commissioning electrical distribution systems
In the simplest terms, commissioning is the process of planning, documenting, scheduling, testing, adjusting, verifying, and training to provide a facility that operates as a fully functional system.
Beneath the umbrella of commissioning are several types, the two largest types being commissioning for new construction and retrocommissioning, which applies the commissioning process to existing facilities that have never been commissioned.
There is a huge difference in approach between these types of commissioning. In new construction, the installing contractors are on-site to assist with the commissioning process as well as perform the actual testing. On a retrocommissioning project, this is usually not the case. You will either need to subcontract a large portion of the work to a contractor who is familiar with these procedures, or you may be able to do some or all of the procedures in-house, assuming you have the properly qualified staff.
When it comes to safety regulations for working with electricity, the water can get so muddy you can walk across it. U.S. Occupational Safety and Health Administration ( OSHA ), NFPA , American National Standards Institute , International Electrical Code , and even the Canadian Standards Organization are all in on the action. There are also several very helpful organizations that provide additional information, including the National Institute for Occupational Safety and Health and IEEE. The important thing to remember is that OSHA safety standards are law. The standards that OSHA refers to are guidelines for safety.
Testing the test instruments
Safety when working with electricity applies not only to how you interact with the systems, but also to how you select your test instruments. Prior to starting any electrical testing, make sure that your test instruments are rated for the voltage ratings of the equipment and systems you will be testing. ANSI C2-81 deals with electric installations of higher than 1,000 V. For voltages under 1,000 V, it is ANSI S82.02 that provides safety rules for electrical test instruments, as well as CSA 22.2-1010.1 and IEC 61010 .
A four-category rating system of over-voltage transient impulses (voltage spikes) was established by these standards. Generally speaking, Category IV systems refer to power lines at the utility connection and the service entrance. This includes outside overhead and buried cable runs.
Category III systems refer to distribution wiring, which includes 460 V 3-phase bus and feeder circuits, distribution panels, and permanent (hard-wired) loads. Category II systems refer to receptacle circuits and plug-in loads. Category I covers protected electronic circuits. Modern test instruments’ category ratings are indicated on the front of the meter, usually near the input terminals. You need to make sure that the test leads you are using are of an equal or higher rating. Meters with these ratings are not only rated for the system voltage, but more importantly are rated for transient-voltage-withstand capability. If your instrumentation is not rated for a minimum of Category III, trash it and buy a rated meter.
Personal protective equipment
Make sure you follow lockout/tag out procedures and wear the necessary personal protective equipment (PPE). This equipment not only helps protect you from accidental electrocution, but more importantly from an arc flash event. The heat generated by the high current flow may melt or vaporize the material and create an arc. This arc flash creates a brilliant flash, intense heat, and a fast moving pressure wave that propels the arcing products. Arc flash temperatures can easily reach or exceed 16,000 F, and the pressure blast can be enormous, enough to knock a person across the room. Other effects of an arcing fault include extreme sound waves, molten metal, and shrapnel.
NFPA Standard 70E and NEC article 110.16 define where and when PPE must be used. Generally speaking, these requirements include different levels of eye and hearing protection, insulated hand tools, insulated gloves, and fire-resistant clothing. The required level of protection is based upon the expected level of incident energy from an arc flash study. There are five rating levels of PPE, from Class 0 (the least amount of protection) to Class 4 (the greatest amount of protection). Lastly, here are a few practices you should adopt that will help you work safely and reduce your chances for an accident:
1 Don’t work alone. Try to have someone else in the general area (not leaning over your shoulder) and be sure they know what to do should something bad happen.
2 Remove jewelry, including your wedding ring. If you’ve been married as long as I have, your finger has probably grown around the ring like a tree around a fence wire. If that is the case, cover your ring with a couple wraps of electrical tape.
3 Do not set your phone on vibrate while working with electricity. Many people are jittery enough just from being near the stuff, and feeling a vibration while you are in the middle of checking voltages can be enough to scare you into jerking your hands into a live connection.
4 Use the one-hand method whenever possible. Set your meter down or hang it somewhere you can see it. Attach the negative or common lead first, and if possible use a lead with an alligator test clip on the end. This leaves one hand free to test with the positive probe, and allows you to keeps your remaining hand in your pocket. The idea is to keep the free hand away from anything metal that can create a path to ground through your body.
Remember, familiarity breeds contempt. Just because you didn’t get electrocuted the last time you did it wrong, that doesn’t mean you will be so lucky the next time.
The main purpose of electric systems commissioning is to increase the reliability of electrical power systems after installation by ensuring a quality installation, identifying problems, and providing a set of baseline values for comparison with subsequent routine tests. As with any commissioning process, a commissioning plan should be developed to provide a planned approach and procedure of what should be done in order to verify the proper system installation. There are numerous commissioning and testing standards that can be referenced, including NEBB , ASHRAE , National Electrical Contractors Assn ., International Electrical Testing Assn. , and numerous others.
Basic testing, instruments, and procedures
Voltage measurements— Voltage measurements are made with a voltmeter, or more often a multimeter that includes a voltmeter function. Used properly, voltmeters are extremely useful instruments. Voltage measurements are very easy to do, yet mistakes are often made. Here’s a surprising fact that most people forget: Voltmeters do not measure voltage. What a voltmeter does is measure the difference in voltage between two points. Try to remember that, because a voltmeter can show 0 V on a circuit that is energized with 460 V. If you’ve ever worked around a colorblind electrician, you already know what I’m talking about. This happens much more frequently in existing buildings where circuits or devices are often added at a later date and power is stolen from an existing circuit.
The biggest mistake is when contractors or commissioning personnel record voltages taken at the equipment disconnect switch with the equipment de-energized. It is correct to make a measurement in this manner prior to starting the equipment, but for testing and recording purposes the voltage must be measured under full load. The correct voltage displayed on the meter with no load confirms that wires are not open, but not that they (or any of the components in the line) will properly support the load. This situation can occur if the circuit has a high-resistance fault, such as when corrosion or rust blocks current flow enough to cause the load (component) to malfunction.
In some cases, full load testing is not always possible to do because of weather, manufacturing processes, critical operations in the facility, etc. In those cases, measurements should be made with the greatest load you can achieve, and data logging can allow monitoring of the circuit so you will not miss the full load measurements. In my area, we often have problems with our 460-V distribution systems. Voltage measurements should never vary more than 10% from the system-rated or nameplate voltage. It is not uncommon for us to measure as much as 510 Vac during early hours in commercial districts. The voltage usually drops to levels of 460 to 480 V once the buildings are up and running, but these periods of high voltage can cause lots of problems with equipment and phase monitors. Unless you are there in the early morning hours, everything will test normal when you are investigating the problem. Additionally, I’ve never had a data logger put in a request for overtime.
Voltage measurements are also necessary to detect the presence of induced and/or backfed voltages. These voltages will be detected by your voltmeter when the circuit should be de-energized. Induced voltages can be caused by adjacent energized conductors located in the same cable tray or conduit, and the amount of induced voltage can be significant.
Backfed voltages are voltages that originate from another circuit or part of the equipment. They can be caused by improper wiring, malfunctioning components, indicating lamps, and transformers. Backfed voltages can be approximately the same value as induced voltages, making determination between the two difficult. Never ground a circuit to determine if the voltage is induced. If you are dealing with a backfed voltage, there is still a generating power source so grounding the conductor will result in arc welding and an unsafe condition. The easiest way to determine if the voltage is backfed or induced is to use a multimeter that includes a low impedance option and a proximity voltage detector. Check the conductor to ground with the low impedance option and check for the presence of voltage with the proximity meter while doing so. If the multimeter and proximity detector both show no voltage, you are dealing with induced voltage. If both the low impedance multimeter and proximity meter show voltage, you are probably dealing with a backfed voltage and the power source needs to be determined and isolated.
Current measurements— Current measurements are also easy to do, but can only be performed on live circuits. For measuring anything but the smallest of currents (such as those found in BAS or fire alarm control systems), the tool that is typically used is the clamp-on meter. Care must be exercised since you are not only working around live circuits, but you must frequently move the energized conductors into a location that allows the meter to be properly used. Interpreting the measurements is straightforward. For power or lighting circuits, you are comparing the measurements to the expected loads on the circuit. For equipment, you are comparing the current readings to nameplate data. You need to be aware of the differences in loads and how system problems will appear.
Resistive loads— These are the easiest types of loads to test. Examples of resistive loads are electric heaters, electric water heaters, incandescent lights, or any other load with a fixed resistance. The actual resistance of these loads will vary slightly as they change temperature, so it is best to energize the load and let it operate for a few minutes prior to testing. A measurement is then made, and the readings compared to the equipment nameplate. This is where the type of load matters. Resistive loads with low voltage will draw low current and loads with high voltage will draw high current when compared to the nameplate.
Inductive loads— Examples of inductive loads include equipment such as motors, compressors, and transformers. The current draw from these loads is related to the load on the device, and in most cases the equipment does not operate at full load. When tested at full load, inductive loads with low voltage will draw higher current and loads with high voltage will draw lower current when compared to the nameplate. Pay particular attention to these measurements, because even though the measured voltage may not be more than 10% higher than the equipment rating, this lowers the full load amperage rating of the motor. For example, if you are commissioning a 25 hp motor that has a full-load amp rating of 28 A at 460-V, 3-phase, 60 Hz but actually measure 26.6 A at 494 Vac, you have exceeded the full load rating of the motor and are now operating in the motor’s service factor. This is why it is important to understand the effects of equipment load and system voltage when making current measurements on inductive loads.
Insulation testing is performed to determine the insulation integrity of the distribution system. This consists of applying a high potential voltage to the item and measuring the leakage current that may flow to ground. There are two types of insulation testing: destructive and nondestructive. Destructive tests are only run one time, usually in the factory to verify the initial strength of the insulation, and nondestructive tests are run as acceptance and maintenance tests to measure deterioration from the original value. Excessive leakage current is an indication of dielectric breakdown and/or impending failure. Insulation may weaken over time at a rate tightly related to the operating time and temperature of operation. For this reason, these tests are usually repeated at scheduled intervals after the system has been turned over to the operating staff to track the insulation deterioration over time.
The instrument used for this test is a MegOhmMeter, also known as a “megger.” These meters measure resistance by inducing a current in the circuit under test, measuring the resultant potential across the circuit, and then calculating the resistance of the circuit. During insulation testing, the circuit must be disconnected from any power source. Meggers typically have several output voltages, usually 250, 500, and 1,000 Vdc. It is considered good practice to test with the lowest voltage first, and then work through the higher voltages with each successful test. Most of the meters on the market today also have a discharge function whereby they will discharge the system being tested through the meter. Before performing any insulation test, you should always consult the component manufacturer for detailed testing instructions and voltage limitations.
Megger testing of conductors can be performed as part of the commissioning process. When performing the test, each conductor is tested with respect to ground and to each adjacent conductor. The test voltage should be 500 Vdc for 300 V-rated cables and 1,000 Vdc for 600 V-rated cables. The test duration is usually one minute. The insulation resistance should not be less than 2 MΩ for circuits under 115 V; 6 MΩ for circuits between 115 V to 600 V with total single conductor length of 2,500 ft or greater; and 8 MΩ for circuits of 115 to 600 V with total single conductor length of less than 2,500 ft.
Megger testing is typically performed on motors greater than 1 hp (746 W). This test is also performed by applying 500 Vdc or 1,000 Vdc directly across the insulation system between the windings and stator frame of the machine (always consult the motor manufacturer for specific test requirements).
NEMA motor manufacturing standards require a minimum resistance to ground at 104 F ambient of one MΩ per rated kV, plus one MΩ. For motors rated 250 Vac or less, the test should be performed at 500 Vdc and the minimum recommended insulation value is 25 MΩ. For motors rated 250 to 600 Vac, the test should be performed at 1,000 Vdc and the minimum recommended insulation value is 100 MΩ. New motors will typically have readings greater than 999 MΩ. Low readings (on any winding) may indicate seriously impaired insulation. This is not always indicative of defective insulation. The resistance value of the dirt, oil, and water that often contaminates the end-winding areas of rotating machineryis quite low and normally results in a high-surface leakage current and a low resistance reading. For this reason, initial testing should be performed on a clean motor, and any scheduled testing by the operating staff must also be performed on a clean motor.
Another good application for megger testing is for heat trace on piping systems. Additionally, the manufacturer of the heat trace will provide its test procedure and required values. In general, the test voltage is usually 1,000 or 2,500 Vdc, and the minimum measured value is typically 20 MΩ. The meter’s negative lead is attached to the metal braid on the heat trace and the positive lead is attached to each conductor.
Thermographic surveys should be performed during periods of maximum loading. Heating is generally related to the square of the current, so the load current has a major impact onT. When conducting the survey, you should record a description of the equipment, list the temperature difference between the component of concern and the reference area, and give the probable cause of the temperature difference. It is also helpful to include load measurements for both the identified component as well as any reference component, and include pictures and thermograms where possible. You will also need to recommend any corrective actions needed.
In general, it is best to compare similar components, but temperature rise above ambient conditions may also be used. A %%TRANGL%%T of 39.2 to 59 F between similar components under similar loading conditions indicates that a deficiency probably exists and the cause should be investigated when the circuit or component can be properly tested. The same is true for aT of 51.8 to 68 F between the component and the ambient temperature. Similar components with similar loading and T of more than 59 or a 104 F T above the ambient air temperature is indicative of a major deficiency, and the circuit or component should be repaired or replaced immediately.
Harmonic analysis studies
In my experience, harmonic analysis studies are not often specified. They are typically required as an additional service when problems with harmonics are identified through other commissioning or start-up tests. Indicators of harmonics include unexplained motor failures, overheating of neutral conductors in balanced loads, transformers overheating when not fully loaded, and circuit breakers tripping when not overloaded. In the simplest terms, harmonics are currents or voltages with frequencies that are integer multiples of the fundamental power frequency. If the fundamental frequency is 60 Hz, the second harmonic would be 120 Hz, and the third would be 180 Hz.
Harmonics are created by nonlinear loads that draw current in abrupt pulses rather than in a smooth sinusoidal manner. These pulses create distorted current wave shapes that in turn result in harmonic currents that flow back into other parts of the power system.
Thermography can indicate the presence of harmonics in the form of overheated neutral conductors, bus bars, and transformers. If harmonics are suspected, there is some limited testing that can be performed with multimeters. If you get significantly different measurements from the two meters on the same phase (15% to 20% or more), it’s a sure sign that you need your meters calibrated or harmonics are present. By measuring the phase with both meters, you are effectively identifying a non-sinusoidal waveform. A frequency other than 60 Hz (neutrals with harmonics will typically read 180 Hz) indicates that harmonics are present. Confirming harmonics can be done with multimeters, but analyzing and mitigating them requires a power quality analyzer.
There are more tests available for commissioning electrical systems than can be discussed in a short article. Hopefully this article has provided insight into some of the testing that can be performed, so that even if you are only observing the tests, you will have a better understanding of what you are witnessing and what to expect.
Huber is a NEBB Certified professional and Certified Energy Manager. As president of Complete Commissioning in Clinton, Md., he has been involved in the installation, testing, and commissioning of electrical and HVAC systems for more than 25 years.