SURGE TEST2016, VALUES AND DIAGNOSTICS FOR ITIG II Surge Comparison .

2016, SURGE TEST VALUES AND DIAGNOSTICS FOR ITIG II Surge Comparison Tests Pass/Fail Recommendations Three Phase AC Stators, Assembled Motors and Generators June 2016 Jacob Beck Kristoffer Torvik Electrom Instruments www.electrominst.com info@electrominst.com 1.720.491.3580

Surge test values and diagnostics for iTIG II Contents Contents 1. 2. 3. 4. Introduction – The Surge Comparison Test 4 1.1. What is a Surge Test used for? 4 1.2. How Surge tests work 4 1.3. Faults Not Found with a Surge Test 6 Pass/Fail Guidelines 7 2.1. Lead to Lead (Phase to Phase) Surge Comparison Tests 7 2.2. Pulse to Pulse (P-P) Surge Tests 7 2.3. P-P Failure Pattern in the bar graph 8 2.4. P-P % between 10% and 30% 9 2.5. Faults not found by a P-P Test 10 2.6. Alternative Test to find welded shorts in an Assembled Motor 10 Case Study and conclusion 11 3.1. Surge Test is non-destructive 11 3.2. Partial Discharge in P-P surge test 11 Appendix – Pass/Fail Guideline Tables 12 2016, Electrom Instruments, Inc. All Rights Reserved. June 2016 2

Surge test values and diagnostics for iTIG II List of Figures and Tables List of Figures and Tables Figure 1.1 Wave Comparison Test. Balanced Phases and Near-Identical Surge Waves . 5 Figure 2.1 Wave Comparison Test. Phase Imbalances - Surge Wave Separation 7 Figure 2.2 P-P Patterns. Failed vs Good. . 8 Figure 2.3 P-P/Wave Comparison. Assembled Motor with Dampened Waves. . 9 Figure 3.1 P-P Pattern. Case study of 5HP, 460V Stator . 11 Table 4.1 Recommendations for Stators . 12 Table 4.2 Recommendations for Assembled Motors . 12 Table 4.3 Recommendations for Pulse to Pulse Test – All Windings/Motors. 12 2016, Electrom Instruments, Inc. All Rights Reserved. June 2016 3

Surge test values and diagnostics for iTIG II Introduction – The Surge Comparison Test 1. Introduction – The Surge Comparison Test 1.1. What is a Surge Test used for? The surge test is used to find shorts or insulation weaknesses in coils, windings, electric motors, generators, alternators and transformers. These faults are typically turn to turn, coil to coil, or phase to phase. They can also be shorts to ground, but IR and Hipot tests will clearly identify these. Other problems that can be found include wrong connections, wrong turn counts and some other issues in DC motor armatures. Weak turn to turn insulation is the source of most winding failures. Because it is the only test that can find weak insulation at elevated voltages, the surge test is important for motor reliability and maintenance programs, for diagnosing problems and for quality control. Surge tests are done on individual coils, stators, wound rotors, and for fully assembled motors and generators without the need to turn the rotor. See Figure 2.3 1.2. How Surge tests work A set of fast rising pulses are passed through the coil or motor (DUT). The voltage of the pulses depends on what voltage the test operator sets, or the standard that is used. The test voltage can range from the RMS operating voltage of the DUT to around 3.5 times the operating voltage of the DUT. 2E 1000V where E is the operating RMS voltage of the DUT is common and recommended by the EASA AR100 standard. The pulses produce a decaying wave in the circuit that includes the tester and the DUT. The wave is compared to the wave from another coil or to the waves from the other motor phases. All are displayed on the screen. See Figure 1.1 The frequency, shape, and amount or ringing of the wave (number of wave peaks) depend on the resistance, inductance and capacitance of the DUT. A 2016, Electrom Instruments, Inc. All Rights Reserved. June 2016 4

Introduction – The Surge Comparison Test Surge test values and diagnostics for iTIG II highly dampened load such as a small assembled motor may only have one peak. Rotors dampen the ringing. See Figure 2.3 Figure 1.1 Wave Comparison Test. Balanced Phases and Near-Identical Surge Waves Good test, all 3 phases overlap, and the WD %’s are very low, 0.3-0.8%. Source: Electrom Instruments, iTIG II Screenshot The waves will be nearly identical if the coils or windings are identical (have the same impedance). If the motor is assembled (the rotor is installed) the rotor may influence the inductance in each phase differently due to transformer action. If so the waves for each phase will differ depending on the angular position of the rotor. See 2.1, P-P surge test. If there is a fault or insulation weakness, the wave for the faulty winding will have a different frequency from the others and the zero crossings will shift to the left. See Figure 2.1. The iTIG II testers calculate and display the difference between pairs of waves/phases (%WD). See 1-2, 2-3 and 3-1 percentages, also referred to as L-L difference (L Lead). 2016, Electrom Instruments, Inc. All Rights Reserved. June 2016 5

Surge test values and diagnostics for iTIG II 1.3. Introduction – The Surge Comparison Test Faults Not Found with a Surge Test There are faults that can only be found with a surge test such as turn to turn insulation weaknesses, but there are also faults that a surge test will not see. These are: Resistive connections internally or externally All strands (round wires) not connected at both ends with several in hand (several in parallel) Wrong gauge wire in a coil or feeder cable Partial Blowout of multi-wire random wound coils where some of the parallel wires are still connected The reason for this is that the surge test results are independent of the resistance in the circuit. Always do winding resistance measurements and comparisons unless this information is available from other sources. 2016, Electrom Instruments, Inc. All Rights Reserved. June 2016 6

Surge test values and diagnostics for iTIG II Pass/Fail Guidelines 2. Pass/Fail Guidelines 2.1. Lead to Lead (Phase to Phase) Surge Comparison Tests The general guideline for a good result is that the L-L %WD should be less than 10%. However, there is a “gray area” around 10% for lap wound stators. Concentric wound stators may have differences up to 20%, sometimes even more. Many assembled motors, especially fast running small motors, may have big differences because of influences by the rotor. The wave difference is 22% in Figure 2.3. The difference can be significantly higher. 2.2. Pulse to Pulse (P-P) Surge Tests This is a test done on one coil, or one phase of a motor. The surge voltage is automatically raised in small voltage steps, and the surge wave at each voltage step is compared to the wave from the previous step. This means that the coil/phase is compared to itself. The %WD is calculated at each step. The maximum %WD seen by the tester is displayed on the screen. A bar graph of all the WD percentages for each voltage step can be displayed by tapping on the P-P % number for each phase. Since the comparisons are to “itself”, the position of the rotor in an assembled motor test does not matter. The tests are independent of other phases or windings. Figure 2.1 Wave Comparison Test. Phase Imbalances - Surge Wave Separation Failed test on L2 (3-1 is good), PP also failed on L2 at 34%. 2016, Electrom Instruments, Inc. All Rights Reserved. June 2016 7

Surge test values and diagnostics for iTIG II Pass/Fail Guidelines Source: Electrom Instruments, iTIG II Screenshot 2.3. P-P Failure Pattern in the bar graph When there is an arc turn to turn in a winding, the bar graph will usually have one or two tall bars above 30% at the point where the arcing started. Figure 2.2 P-P Patterns. Failed vs Good. Failed P-P test. The large bar indicates a flashover. Good P-P test – low P-P%. No large spikes in graph. The P-P test is very useful for testing the following: Assembled motors: (rotor installed) where the rotor position influences the surge waves. Concentric wound motors: They sometimes have a large separation of the surge waves indicating that there might be a turn to turn fault (arc). The P-P test will determine if there is an arc or not. Single phase motors or other coils/windings where there is no other winding to compare to. Any time a phase to phase comparison is questionable. Source: Electrom Instruments, iTIG II Screenshot 2016, Electrom Instruments, Inc. All Rights Reserved. June 2016 8

Surge test values and diagnostics for iTIG II 2.4. Pass/Fail Guidelines P-P % between 10% and 30% Pass/fail interpretation in two common cases of elevated levels of P-P: If there are a few bars at the beginning of the graph (left side - low voltages) up to 30%, the result is a pass. If there are many bars on the right side of the graph (high voltages) that are up to 30%, Partial Discharges (PD) have been detected. See Figure 3.1. Flash-over in weak insulation (arcs) is represented by one or two bars above 30%. NOTE: The P-P test is not designed to be a PD measurement and does not accurately detect when PD starts. But, with enough PD several higher bars will usually be present in the P-P bar graph. The presence of PD does not necessarily mean there is a failure or problem, but can mean that the insulation is starting to weaken, especially in low voltage motors. Around 30% and above, the PD typically turns into an arc. Figure 2.3 P-P/Wave Comparison. Assembled Motor with Dampened Waves. Test Passed, L-L% is high, but P-P% is below the limit. Source: Electrom Instruments, iTIG II Screenshot 2016, Electrom Instruments, Inc. All Rights Reserved. June 2016 9

Surge test values and diagnostics for iTIG II 2.5. Pass/Fail Guidelines Faults not found by a P-P Test The P-P test looks for a change in the waves as the surge voltage rises. Therefore, if there is a hard short, or welded short turn to turn in a winding, and at the same time the insulation around this short is stable and strong, the P-P test will not find it since there will be no change as the surge voltage rises. This fault can be seen with an L-L surge test. But, if the motor is assembled, one cannot usually tell if it is rotor influence or a fault unless the fault is severe. High precision resistance measurements may tell if windings have shorted turns. If the number of turns measured are more than say 50 or 100, this also becomes harder to interpret since the welded adjacent turns change the resistance by an amount that may be less than the failure limit. An Inductance measurement can also be used, but not in an assembled motor with rotor influence. 2.6. Alternative Test to find welded shorts in an Assembled Motor If it is possible to turn the rotor between surge tests of the 3 phases, this can help find a welded short. The rotor must be turned between tests so that the waves line up. If they do line up there is no problem. If the minimum difference attained is significant, there is a problem. Beware of potential rotor problems when doing this test. 2016, Electrom Instruments, Inc. All Rights Reserved. June 2016 10

Surge test values and diagnostics for iTIG II Case Study and conclusion 3. Case Study and conclusion A 5HP 460V stator from 1980 was taken out of service around 2005. This stator remained in storage for ten years. In 2015 it was cleaned and blown out with high a pressure air and used in a case study. The stator first arced when the surge test voltage reached 8000V. It has since been tested repeatedly and the arc voltage remains in the 7500V to 8000V range. 3.1. Surge Test is non-destructive This is a good example of why surge tests are not destructive when done properly. A 460V motor like this is typically tested to 1920V, not 8kV. The stator withstand strength remains unchanged even after arcing. 3.2. Partial Discharge in P-P surge test Partial Discharge starts a little below the test voltage, and can be seen in this picture as the group of taller bars 10% to 20% high towards the high end of the graph. Note that the L-L %WD is still in the good range. Figure 3.1 P-P Pattern. Case study of 5HP, 460V Stator Source: Electrom Instruments 2016, Electrom Instruments, Inc. All Rights Reserved. June 2016 11