Rotor Installed Corona Mapping of Stator Windings within Large Diameter Hydro Generators
Previous work has shown that existing technology (near field communication (NFC) antennas) from other industries (consumer electronics and the internet-of-things) can be leveraged to obtain partial discharge measurements of a stator coil from within a simulated air gap. With this knowledge, three questions remain. First, can reliable measurements be taken from within actual machines, over entire stator windings, with localization resolution to a specific coil or coil group, without removing the rotor? Secondly, if this technology can be leveraged in actual machines, can a robust prototype be developed which can facilitate on-demand deployment for acquiring such data across Reclamation's rotating machines? And finally, if a reliable prototype can be developed and backed by several field trials, can a technology transfer be completed in order to further develop the prototype toward the largest possible benefit within the power industry?
If this research is funded, these three research questions would directly correlate to three research phases to be carried out pending the results of its preceding tasks. Specifically, phase 1 intends to show the results obtained in the lab in previous research can be replicated in the field in real rotating machines. Moreover, Phase 1 also intends to make a direct comparison of using NFC antennas to map corona when compared to the present method of corona probe testing. This will be carried out through research partners which will provide a machine for testing with the rotor removed. A normal corona map using present corona probe testing techniques will be carried out. In addition, an apparatus will be made to be affixed to the end of a hot-stick which will position an NFC antenna in front of the stator core as if it was affixed to a rotor pole face. This apparatus will be moved about the stator winding face with data collected similar to the corona probe. Data analysis will then be carried out to determine a correlation, if any, b
Need and Benefit
Presently, partial discharge detection in rotating machine high-voltage insulation systems is of two common methods:
1.Capacitive couplers, galvanically connected to the winding, usually at the high-voltage end of the winding.
2.Corona probe testing, which is a bulky inductive antenna at the end of a hot-stick, which, with the rotor removed, is
moved around the energized winding by the test technician. The measured magnitude and location of measurement
are then notated by hand for later analysis.
The capacitive coupler method can trend data over years, however, it cannot localize measurements within the
winding, or measure more than a small fraction of the entire winding. While the corona probe method can localize
measurements to specific locations within the winding, this method is rarely performed due to the associated costs
and risks, and therefore is not trended. In short, present technology can either trend but not localize, or localize but
not trend—either way, a sacrifice in diagnostics must be made. This sacrifice is significant because the absolute
magnitude of measured partial discharge differs from system to system. In other words, a relatively high reading on
one machine may be cause for concern, while the same reading on a different machine may not. This is why the data
for a given machine must be trended, which would produce concern when a significant increase in activity from
previous results is observed. Similarly, without the ability to locate the source of the discharge activity, it is impossible
to gauge the severity of the deterioration. More specifically, without localization, it cannot be determined whether the
discharge activity is a bulk issue (in which case the entire machine must be rewound with significant cost), or an issue
in a small section of the machine (in which case one or more coils can be replaced to get more life out of the asset at
much less cost).
Many rotating machines within Reclamation's fleet can cost several thousand to several hundred thousand dollars a
day during an unplanned outage. This loss can be in the form of lost revenue, costs associated with purchasing
replacement power for lost generation, and inability to meet water delivery agreements. For this reason, it is
imperative that unplanned outages are mitigated. This can be achieved with more accurate diagnostics and trending.
Specifically, if asset managers can trend insulation deterioration mechanisms, while also obtaining a picture of where
the deterioration is occurring within the machine, they can better allocate funds and plan machine replacement or
repair. Without accurate diagnostics and trending, the probability of unexpected failure and outages increases. Such
unexpected failures can often lead to extended unplanned outages while funding for repair or replacement is secured,
exacerbating overall costs. These costs, which are significant throughout the industry, set the need and urgency for
bringing better diagnostic testing and trending technology to market. The proposed technology hopes to advance this
Contact the Principal Investigator for information about partners.
Contact the Principal Investigator for information about these documents.