| |
|
|
|
|
|
| |
|
|
|
|
|
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
|
The Doble SFRA Resource CenterFAQ's Q: What are FRA and SFRA? A: FRA stands for Frequency Response Analysis which is diagnostic test for power transformers; SFRA stands for Sweep FRA to differentiate the Doble M5100 test set from less reliable impulse based devices. Q: Why is SFRA important? A: SFRA finds mechanical problems in transformers (winding movement and/or distortion) caused by events such as electrical faults or after transportation of the transformer. Q: Why bother? A: If your transformers look after themselves, don't suffer and never need to be moved, well... we'd like to meet you. BUT - If you ever need to make critical decisions about the suitability of a transformer for service, or check on integrity after relocation, then SFRA is an indispensable tool. Q: How long does it take to do the test? A: A couple of hours for a large transformer. Usually block out half a day for testing with SFRA. Q: Aren't Impulse Systems and the M5100 doing the same thing? A: They're making the same measurement, but doing it in different ways: the M5100 sweeps through the frequency spectrum in a consistent and reliable way making measurements at each individual point, impulse systems try to do all frequency measurements at once and fail. Q: Why are impulse systems poor? A: Because they do not have a repeatable enough impulse to provide a good signal; the impulse is contaminated by noise; commercial systems are very sensitive to lead movement on repeat testing; data acquisition is compromised by a need to measure millivolts simultaneously with signals of tens of volts. It would take a lot more money to produce a repeatable and reliable impulse based system for field use. Q: Why is sweep better? A: Because a measurement is made at each frequency point; the signal at a particular frequency is output for a number of cycles to allow the electronics time to settle; a number of cycles are measured at each frequency and noise eliminated through signal processing; gain stages are used to measure the response amplitude and provide maximum resolution of the measured response Q: What's does a typical result look like and what are the main features? A: The response varies for transformer size, winding configuration and winding under test. Typical results are available. An overall comparison of different results - LV, HV, Interwinding and Short Circuit helps differentiate between results. Q: What do the resonances mean? A: The resonances at a given frequency relate to a combination of inductance, L, and capacitance, C, within the transformer. The transformer is a complex web of inductances, mutual inductances and capacitances between windings and to ground. Consequently there are many resonances in the SFRA plot. The L's and C's within a transformer are strongly related to geometry, hence the sensitivity of the SFRA test to movements within a transformer. Q: Are there any general rules for what the resonances mean? A: Yes. Generally, the lower frequencies relate to larger objects and higher frequencies relate to smaller objects within the transformer. Smaller transformers usually have more higher frequency responses. The first low frequency resonance relates to core, followed by resonances relating to overall windings, then winding sections, tap windings and smaller structures within the transformer. Q: Can we use sister units as a reference? A: Yes. Manufacturing tolerances and design changes influence results, so interpreting results with reference to a sister unit is never as good as with reference to a baseline. A technical paper on use of sister units discusses these issues in detail and gives some good examples. Q: What do shifts in frequency mean? A: Resonance shifts to the left or right - frequency shifts - mean something has happened to the circuit L or C. This is probably significant: something has changed the geometry of the L's and/or C's. Where a resonance reduces or increases in amplitude, this relates primarily to the R, and may well be related to contact resistance; however, L and C variations, particularly at higher frequencies are still possible with contact impedance. Basics 1 and Basics 2 look at these variations in more detail. Q: Can we automate the process of analysis? A: To a degree, but experience has shown that any rules for analysis always have exceptions. SFRA provides a simple picture and reducing that to a number loses huge amounts of information. Doble's software provides simple difference plots so as to identify areas of difference in the two traces. Q: Why are good connections important? A: Electrical tests of any type should be performed with attention to good test practices. The SFRA is a very powerful and sensitive test which will be slightly affected by resistance in measurement or ground clips. Q: What if there are new or missing resonances? A: This is strong evidence that there has been substantial winding movement and/or distortion within the transformer. Cases of false positives - new peaks being due to the test set - have occurred with impulse systems. This means that a reliable assessment of transformer condition is very difficult. Q: Is it possible to test a transformer without oil? A: Yes. Testing without oil provides different results as the dielectric constants within the transformer have been changed as air has replaced oil. Suitable precautions must be taken to ensure safety when testing without oil. Q: Do results change with tap position? A: It is important to record the tap position as part of the test records as results do vary with tap position. This is a well known effect and good procedures are useful in making sure that this information is recorded. In addition, when testing at neutral, the position of the reversing switch may also affect results so it is necessary to record whether neutral was arrived at from a raised or a lowered position. Q: How important is grounding? A: As SFRA is an electrical test, good grounding is required for repeatable and reliable results. Q: How does core magnetization affect results? A: Core magnetization is a well understood low frequency effect and can be identified in results. Demagnetizing the core removes the effect but is not necessary when making an assessment. Q: What measurements should we make? A: A connection guide is available which covers most of the popular transformer designs. In general, windings are tested separately in as small a section as possible: HV, LV, Tertiary are tested separately and phase by phase. This means that, in future, possible problems may be located more precisely. Usually other windings and phases are left open circuit during a test, except in the case of short circuit SFRA measurements. Q: What are short circuit SFRA measurements? A: Short circuit SFRA measurements are a useful addition to standard open circuit tests. The short circuit test removes the effect of the core at low frequencies and allows for direct phase-to-phase comparison on a transformer without reference results. Q: What standards are being developed for SFRA? A: There are two independent organizations that are looking at standardizing FRA activity: the IEEE and CIGRE. Doble contributes to both efforts through written contributions and meeting attendance; we promote FRA measurements which are repeatable, reliable and useful. Q: How can I include FRA in my transformer purchase specification? A: Doble provides a generic Transformer Purchase Specification which is not limited to sweep systems and is not limited to Doble. It does require that the measurements made are repeatable, and reliable by insisting on measurement criteria appropriate for FRA testing. For the latest version, contact Doble. Q: Question still not answered? A: Please contact us at SFRA_UG@doble.com and we will do our best to address your concerns, or write to us at: SFRA Product Manager Office phone: ++1
(617) 926 4900
|
|
