SUMMARY OF TOPICS



The following is a brief summary of the new paragraphs in the Standard for Electric Generators, UL 1004-4, which have a future Effective Date of September 15, 2010.

Subject 2520 (1004-4)

SUMMARY OF TOPICS

The following changes in requirements are being proposed:

1) The proposed first edition of UL 1004-4, Electric Generators. There are no new technical changes being proposed, just the separation of electric generator requirements into a separate standard.

STP BALLOTS & COMMENTS DUE: JANUARY 15, 2008

Reviewed by:

Copyright © 2007 Underwriters Laboratories Inc.

1. The proposed first edition of UL 1004-4, Electric Generators. There are no new technical changes being proposed, just the separation of electric generator requirements into a separate standard.

RATIONALE

Proposal submitted by Frank Ladonne, UL

UL proposes the first edition of UL 1004-4, Electric Generators. The proposed UL 1004-4 is a standard which is part of a series of standards. UL 1004-1, Rotating Electrical Machines - General Requirements, addresses the general requirements for all rotating machinery, and subsequent standards in the series, such as UL 1004-4, address the particular product requirements and cover specific constructions. When the combination of UL 1004-4 and UL 1004-1 are compared with its predecessor, Subject 1004B, Outline of Investigation for Electric Generators, the combination of UL 1004-4 and UL 1004-1 proposes no requirements that are new or unique.

The UL 1004 series is a consolidation and rewrite of UL 1004, 1004A, UL 1004B, and UL 2111. The rewrite and consolidation of the rotating machinery requirements is intended to result in Standards and requirements that are more reflective of current and emerging technologies such as brushless DC (BLDC) or electrically commutated motors (ECM), servo motors, stepper motors, and the like. In addition, this is intended to result in Standards that represent the most current technical philosophies. UL 1004, UL 1004A, UL 1004B, and UL 2111 will eventually be withdrawn after the new UL 1004-1 and series standards are published.

Section 41D, Surge Tests, was added as acceptance criteria for the short-circuit test. The pulse-to-pulse EAR test is included as yielding quantifiable results while being non-destructive as opposed to the dielectric test that has typically been used as acceptance criteria. Surge testing is gaining increasing use in the industry and has been included in other prominent rotating machinery Standards including ANSI/NEMA MG-1, Standard for Motors and Generators, and IEEE 522, Guide for Testing Turn Insulation of Form-Wound Stator Coils for Alternating-Current Electric Machines.

[pic]

INTRODUCTION

 

1 Scope

 

1.1 This Standard is intended to be read together with the Standard for Rotating Electrical Machines - General Requirements, UL 1004-1. The numbering of the Sections in this Standard corresponds to the like numbered Sections in UL 1004-1. The numbering of the paragraphs are unique to this Standard, and do not directly correspond to the like numbered paragraphs in UL 1004-1. The requirements in this Standard supplement or amend the requirements in UL 1004-1. The requirements of UL 1004-1 apply unless modified by this Standard. For Sections not shown, refer to the Standard for Rotating Electrical Machines - General Requirements, UL 1004-1.

 

1.2 This Standard covers electric generators, sometimes referred to as generator heads, which, when coupled with prime movers, such as engines or electric motors, are used to produce electricity. This Standard covers generators (DC machines) and alternators (AC machines) rated 7,200 volts or less.

 

1.3 This Standard does not cover portable or stationary generator assemblies, which are respectively covered under the Standard for Stationary Engine Generator Assemblies, UL 2200, and the Outline for Portable Engine-Generator Assemblies, Subject 2201.

 

1.4 This Standard does not cover generators intended for use in hazardous locations as defined in the National Electrical Code ®, NFPA 70.

The National Electrical Code® and NEC® are registered trademarks of the National Fire Protection Association, Inc., Quincy, MA 02169.

 

1.5 In the context of this document, the term generators shall be understood to apply to both AC and DC machines.

 

2 Components

 

2.1 Voltage regulators shall comply with the requirements of the Standard for Industrial Control Equipment, UL 508.

PERFORMANCE

 

41A Overspeed Test

 

41A.1 The generator, arranged for operation in its intended manner, shall be subjected to an overspeed condition as described in 41A.2.

 

41A.2 The generator is to be connected to a resistive load, run at rated RPM and then the load adjusted so that the generator is delivering full rated output power. The rotational velocity of the generator is then to be increased to 120 percent of rated RPM and maintained at that speed for 1 minute.

 

41A.3 There shall be no evidence of a risk of fire, shock, or personal injury as a result of this test.

 

41B Short Circuit Test

 

41B.1 A generator provided with an output overcurrent protective device shall be subjected to the test described in 41B.2.

 

41B.2 The generator is to be run at rated RPM and then the output shall be short-circuited. For multi-phase machines, each phase shall be individually and separately short-circuited both phase to phase as well as phase to ground (or neutral).

 

41B.3 As a result of this test, the overcurrent protective device shall operate as intended and there shall be no evidence of a risk of fire, shock, or personal injury.

 

41B.4 At the conclusion of the Short Circuit Test, the generator shall be subjected to and comply with Section 41D, Surge Tests.

 

41C Output Waveform Distortion

 

41C.1 When tested as described in 41C.2, the total rms value of the harmonic voltages, excluding the fundamental, and the rms voltage of any single harmonic, delivered by a generator shall not exceed the values in Table 41C.1.

 

41C.2 The generator is to be connected to a linear load having an adjustable impedance so that the generator is able to deliver power at rated power factor. The measurements are to be conducted at open circuit (0 percent), and with the generator delivering 33 percent, 66 percent, and 100 percent of rated power.

Table 41C.1

Generator output - RMS distortion limits

|Harmonic  |15KW or Larger Generator Distortion Limit (Percent)  |Less than 15KW Generator Distortion Limit (Percent)  |

|Odd  |  |  |

|3rd through 9th |4.0  |8.0  |

|11th through 15th |2.0  |4.0  |

|17th through 21st |1.5  |3.0  |

|23rd through 33rd |0.6  |1.2  |

|35th through 39th |0.3  |0.6  |

|Even  |  |  |

|2nd through 10th |1.0  |2.0  |

|12th through 16th |0.5  |1.0  |

|18th through 22nd |0.375  |0.75  |

|Total Harmonic Distortion |5.0 |10 |

 

41D Surge Tests

 

41D.1 At the conclusion of the Short Circuit Test of Section 41B, the generator shall be subjected to and comply with the Surge Test of this Section (see Appendix A). The Surge Test is to be conducted at an amplitude of 3.5 pu, where 1 pu is defined as:

[pic]

in which:

     VL is the rated rms line-to-line voltage

and with a pulse rise time between 100 ns and 1.2 μs.

 

41D.2 Pulse-to-Pulse Error Area Ratio (EAR) Test - Each phase of the generator under test shall be subjected to a series of impulses with each impulse having the rise time described in 41D.1. Each impulse is to be the result of a capacitor discharge. The pulses are to be applied with increasing amplitude from zero to VL with a difference of no more than 25 volts between successive pulses. The EAR between successive pulses (defined in the equation below) is to differ by no more than 10 percent. The formula for EAR between successive pulses is given by:

[pic]

in which:

Fi(1) is a point in the time series of the first waveform;

Fj(2) is the corresponding point in the time series of the second waveform;

Npts is the number of points in the time series sampled at each "jth" point; and

EAR1-2 is the Error Area Ratio of the test Fj(2) waveform with respect to the reference Fi(1) waveform.

 

[pic]

Appendix A - Surge testing of induction machines

 

A1 Principle of the Surge Test

 

A1.1 If a rapidly increasing current is applied to a coil, a voltage will be generated across the coil by the principle of induction. The voltage across the coil is given by V=L*di/dt

in which:

     V is the terminal voltage across the coil;

     L is the coil’s inductance; and

     di/dt is the time derivative of the applied current pulse.

 

A1.2 The terminal voltage V at the leads of the coil is actually a summation of the induced voltage created between individual loops in the coil. If the insulation separating adjacent coils is weak and if the induced voltage is higher than the dielectric strength of the weak insulation, an arc will form between the coils. Surge testing equipment is designed to create the induced voltage between adjacent coils and detect the arcing indicative of weak or failing insulation.

 

A1.3 Figure A1.1 shows a block diagram typical of today’s instrumentation. The internal capacitor is charged to a known voltage by the power supply. At a specific time, a high voltage switch closes which transfers the charge from the capacitor through the windings of the coil. If the resistances and loss of the entire circuit are such that the system is under damped, charge will be able to flow through the inductor and on to the other side of the capacitor resulting in an oscillation This process of ringing will repeat until the resistances and losses in the circuit completely absorb all of the energy that was originally on the capacitor. The measurement of the terminal voltage of the coil vs. time gives the surge waveform, which shows the damped oscillation.

 

Figure A1.1

Block diagram typical surge instrumentation

[pic]

 

A1.4 The ringing frequency of the dampened sinusoidal waveform will be according to the following formula:

[pic]

in which:

f = the ringing frequency of the resulting waveform

L = the inductance of the machine winding under test

C = the capacitance of the internal charge capacitor

R = the system resistance

If the turn-turn insulation fails with an arcing short between two turns in the coil, a fraction of the inductance will be shorted out of the circuit. From the equation above, the ringing frequency f will increase as the inductance decreases due to the short. An increase in the ringing frequency will show itself to be a jump to the left of the ringing pattern. To reiterate, it is this sudden increase in ringing frequency that is the indication of the arcing turn-turn fault. Depending on the coil and the location of an arcing short, the magnitude of the surge waveform may also slightly decrease. Today’s instrumentation will slowly increase the test voltage and "look" for the increase in ringing frequency.

 

A2 Automatic Fault Detection Methods

 

A2.1 The greatest advancement in surge testing has come about with use of high-speed analog and digital electronics and the application of computers to control the testing process. Algorithms programmed into the computer can detect small variations in the shape of the waveform that people’s eyes miss. An additional benefit to having a computer control the test is the immediate shutdown of the test after the insulation "fails" a single pulse. The detection algorithms include:

a)     Zero Crossing Shift

b)     Pulse-to-Pulse EAR

c)     Line-to-Line EAR

 

A2.2 Pulse-to-Pulse EAR - Where the computer has the advantage over the human is detecting slight changes of the waveform shape. The computer uses the Error Area Ratio (EAR) to get a quantitative measure of the difference in shape of two subsequent waveforms called the Pulse-to-Pulse EAR (ppEAR). Before describing the ppEAR, the EAR calculation will be introduced.

The formula for the Error Area Ratio is:

[pic]

in which:

Fi(1) is a point in the time series of the first waveform;

Fj(2) is the corresponding point in the time series of the second waveform;

Npts is the number of points in the time series sampled at each "jth" point; and

EAR1-2 is the Error Area Ratio of the test Fj(2) waveform with respect to the reference Fi(1) waveform.

This simple formula is a fast calculation for a computer, yet it is very accurate at detecting a difference in the shape of the two waveforms. Two exactly identical waveforms will have an EAR of 0%. Two waveforms that look identical to the eye can have EAR values in the 4-5% range. Two waveforms with a noticeable difference will have an EAR value in the range of 10% or higher.

The application of this formula for the ppEAR is to compare the most recently acquired waveform to the previously acquired waveform as the test voltage is slowly increased. The two waveforms are expected to be different since one is at a slightly higher test voltage than the other. This difference is on the order of 4%. However, if the most recently acquired waveform is above the arcing voltage for the failed insulation, its shape will change. The change in shape will then be detected by the EAR algorithm.

Figure A1.2 shows the ppEAR detecting an arcing turn-turn insulation failure.

 

Figure A1.2

Example of coil failing pulse-to-pulse EAR

[pic]

Figure A1.2 shows two surge waveforms and a small ppEAR graph in the lower right corner of the display. The two waveforms represent the previous to fail and the failed waveform. The small graph shows the running ppEAR where the first value plotted is for the two pulses at 800V and 810V. The ppEAR values for successive pulses, each 10 volts higher than the previous pulse, are shown along the majority of the graph until a pulse is applied above the dielectric strength (|P52700V) for the weak insulation. The ppEAR for the failed pulse compared to the previous to fail pulse is shown completely off the scale on the graph. The dashed line along the top of the graph represents the user defined maximum allowable ppEAR for the test. Anything above this value will fail the test. In the figure above, the maximum allowable ppEAR was set to 5% and the "failed" ppEAR was 12%.

The ability to perform such detection schemes as the ppEAR is a testament to the advancement of the surge testing instrumentation available today. The ability to precisely control the impulse voltage, to precisely trigger the impulse generator, and the ability to implement advanced signal processing algorithms during the testing process are key to being able to find the very small changes in surge waveforms that represent insulation failures.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download