Mike Howell
EASA Technical Support Specialist
The question of surge test voltage levels used for testing the turns insulation of form wound stators comes up often. And, why wouldn’t it? Many service centers have several good coil suppliers in their qualified supplier lists and quite a few customers who own form wound motors. It’s not unlikely that each of the coil suppliers and each of the customers could be specifying different test levels. What’s more, most authoritative standards or guides available provide a range of test levels. This is because not all insulation systems are designed to have the same surge withstand capability. Also, the turn insulation properties of one particular system can vary widely depending on the extent of processing complete at the time of test.
Since the vast majority of form wound stators produced today are global vacuum pressure impregnation (VPI) processed, a common challenge is testing at a high enough voltage to find significant defects before processing and at a low enough voltage to avoid damage to the “green” insulation due to its unfinished state. Work done in this area has shown that testing with a floating core (not grounded) can reduce the electrical stress to ground by as much as 50% when testing coils in the stator before VPI processing.
Surge amplitude and rise time
Before discussing when to surge test and at what specific test levels, let’s review the two terms most commonly used to describe the test. Figure 1 shows a typical surge waveform with time plotted on the horizontal axis and voltage on the vertical axis. The two values noted are the surge amplitude and the rise time. The surge amplitude here is referring to the first peak of the waveform, and this is what we normally call the test voltage. The time it takes for the waveform to rise from 10% to 90% of the surge amplitude is called the rise time (IEEE 522), and the surge will reach the peak value at 125% of the rise time. Note that IEC 60034-15 defines the front time as the time to reach the peak value. In order for the test to stress the turn insulation, the rise time must be sufficiently small (well below 1 microsecond). Typically employed standards require a minimum of 3-5 surges and most commercial equipment far exceeds this count in the few seconds needed to review the waveform.
The surge withstand capability of the winding should be verified at one or more of the following steps of manufacture: (a) individual coils before installation in slots, (b) individual coils after installation but prior to connection, with wedging and bracing in place, and (c) on the completely wound and fully-cured stator. Best practice is to do testing at steps (a) and (b) and an additional reduced voltage test after connection but before processing. The reduced voltage test can provide some assurance that the connection is correct.
The test levels should be agreed upon in advance by the coil manufacturer, service center, and in some cases, the customer. Depending on the application and design of the insulation system, typical values for fully-cured coils (100%) and “green” or uncured coils (60%) are shown in Table 1 on Page 4 for some common machine voltages. The test values shown are consistent with IEEE 522 and IEC 60034-15 though the standards include reduced voltage levels ranging from 40% to 80%. IEEE 522 refers to 3.5 per unit (p.u.) as a standard withstand voltage and 2.0 p.u. as a reduced voltage test used for windings that are not likely to see high-magnitude, fast-fronted surges (where 1 p.u. = peak volts to ground of stator winding). NEMA MG-1 takes the opposite approach, referring to 2.0 p.u. as the standard test voltage and 3.5 p.u. as a test level for windings designed for higher surge capabilities. IEC 60034-15 defines the interturn insulation test voltage, U'p as a function of the rated peak lightning-impulse withstand voltage, Up.
We can look at a couple of examples of how the surge test values might be calculated for different scenarios.
Example 1
A 4 kV stator is being rewound for global VPI processing and the customer’s PO states that the winding should be designed for surge withstand capability in accordance with IEC 60034-15.
Up = 4 VL-L + 5 kV = 4 · 4 + 5 = 21 kV
U’p = 0.65 Up = 0.65 · 21 = 13.7 kV
For compliance with IEC 60034-15, surge testing performed prior to VPI could be done at a reduced voltage in the range of 5.5 kV (40% of 13.7 kV) to 11.0 kV (80% of 13.7 kV). If the coil manufacturer determines that 5.5 kV is too high for the materials being used, either the insulation system would have to be modified or an exception to the customer PO would be needed. If a surge test was required on the finished winding after VPI, it would be done at 13.7 kV.
Example 2
A 2.3 kV generator stator is being rewound in the field using dip & bake style coils. These coils have fully-cured turn and ground insulation. The rewind specification requires a surge withstand capability of 3.5 per unit in accordance with IEEE 522.
1 p.u. = 2.3 √(2/3) = 1.88 kV
3.5 p.u. = 3.5 · 1.88 = 6.6 kV
If designed and manufactured properly, these fully-cured coils could safely be surge tested at 6.6 kV in the coil shop after manufacturing and once again after wedging and bracing. Repeating the factory coil test after wedging and bracing ensures that the installation process did not cause or reveal a turn-to-turn short. Surge testing of completed form windings can help identify connection errors, but it is not very effective for testing turn insulation of the complete winding.
The service center should take great care to understand the insulation systems they are using in form wound stators to prevent in-process damage to insulation. It is a good idea to periodically review what you are specifying with your coil suppliers and to always review customer specifications prior to coil manufacturing. With some insulation systems intended for global VPI, even the 60% levels shown in Table 1 may be too high for in-process tests. Well-constructed “green” coils may still pass tests at excessive test levels, but partial discharge activity occurring in air pockets between tape layers could initiate tracking that would reduce the life of the finished winding. As with many special processes, you cannot ensure quality by testing it in. You have to build it in with qualified personnel, quality materials and controlled processes. With the appropriate test levels, dependent on the insulation used, significant defects should be detectable though, even at reduced voltages. Take care to avoid “over-testing” windings. Too many subsequent tests at full test voltage can damage an otherwise good winding.
AVAILABLE IN SPANISH
ANSI/EASA AR100
More information on this topic can be found in ANSI/EASA AR100
Print