Methods and systems for enhancing control of power plant generating units

Patent metadata

Field	Value
Publication number	US-2016147204-A1
Application number	US-201414555221-A
Country	US
Kind code	A1
Filing date	Nov 26, 2014
Priority date	Nov 26, 2014
Publication date	May 26, 2016
Grant date	—

Abstract

A method for operating a sensor in a thermal generating unit. The method may include the steps of: defining lookback periods, wherein the lookback periods each include previous periods of operation for the thermal generating unit, the lookback periods including at least a first lookback period and a second lookback period; receiving a first dataset regarding readings for the sensor during the first lookback period; receiving a second dataset regarding the readings the sensor during the second lookback period; performing a first check on the first dataset and obtaining therefrom a first result; performing a second check on the second dataset and obtaining therefrom a second result; and determining a likelihood as to whether the sensor is malfunctioning based on the first and the second results.

First claim

We claim:

1 . A method for operating a sensor in a thermal generating unit, wherein the sensor is communicatively linked to a control system and configured to take readings so to measure an operating parameter related to an operation of the thermal generating unit, the method comprising the steps of:
defining lookback periods, wherein the lookback periods each comprise previous periods of operation for the thermal generating unit, the lookback periods including at least a first lookback period and a second lookback period; receiving a first dataset regarding the readings for the sensor during the first lookback period; receiving a second dataset regarding the readings for the sensor during the second lookback period; performing a first check on the first dataset and obtaining therefrom a first result; performing a second check on the second dataset and obtaining therefrom a second result; and determining a likelihood as to whether the sensor is malfunctioning based on the first and the second results.

2 . The method according to claim 1 , wherein the first lookback period comprises a short lookback period, and the second lookback period comprises a long lookback period, wherein the second lookback period is multiple times longer in length than the first lookback period.

3 . The method according to claim 1 , wherein the first lookback period comprises a short lookback period of approximately several minutes, and the second lookback period comprises a long lookback period of approximately several hours.

4 . The method according to claim 3 , wherein the first lookback period comprises approximately 1 to 10 minutes and the second lookback period comprises approximately 1 to 3 hours.

5 . The method according to claim 1 , wherein the second lookback period comprises several regularly spaced intervals, each of the intervals comprising an approximate same length as the first lookback period; and
wherein the first lookback period comprises a latest one of the intervals of the second lookback period.

6 . The method according to claim 5 , wherein a number of the intervals included in the second lookback period comprise between approximately 10 and 20.

7 . The method according to claim 5 , wherein the second check comprises a model check the includes the steps of:
calculating predicted values that correspond to measured values of the readings of the second dataset; and comparing the predicted values against the corresponding measured values from the second dataset; wherein the predicted values are derived from a simulation of the operation of the thermal generating unit.

8 . The method according to claim 7 , wherein the first check comprises a continuity check that includes the steps of:
determining whether a total number of the readings included in the first dataset is greater than a minimum allowable threshold; and determining a percentage of the total number of readings that comprises non-available readings, and then determining if the percentage is less than a maximum allowable threshold.

9 . The method according to claim 7 , wherein the first check comprises a range check that includes the steps of:
defining a range between a maximum threshold and a minimum threshold, wherein the range is based upon values of historic readings of the sensor; determining whether the readings included in the first dataset comprise values within the defined range.

10 . The method according to claim 7 , wherein the first check comprises an averaging check that includes the steps of:
calculate average values for the readings in the first dataset, wherein the average value comprises the averaging of corresponding readings from the sensor and at least one other sensor of the same type; defining a range about the calculated average values in which: a positive offset from the calculated average values comprises a maximum threshold; and a negative offset from the calculated average values comprises a minimum threshold; determining whether the readings included in the first dataset comprise values within the defined range.

11 . The method according to claim 7 , wherein the first check comprises a data check that includes determining whether a sequential plot of the readings of the first dataset over the first lookback period comprises a profile indicative of a data irregularity.

12 . The method according to claim 11 , wherein the data irregularity comprises the profile showing a data shift in the sequential plot of the readings.

13 . The method according to claim 11 , wherein the data irregularity comprises the profile showing a data drift in the sequential plot of the readings.

14 . The method according to claim 11 , wherein the data irregularity comprises the profile showing a data spikes in the sequential plot of the readings.

15 . The method according to claim 11 , wherein the data irregularity comprises the profile showing at least one of increasing noise, decreasing noise, and senility in the sequential plot of the readings.

16 . The method according to claim 7 , wherein the first check comprises a continuity check that includes the steps of: determining whether a total number of the readings included in the first dataset is greater than a minimum allowable threshold; and determining a percentage of the total number of readings that comprises non-available readings, and then determining if the percentage is less than a maximum allowable threshold;
wherein the first check comprises a range check that includes the steps of: defining a first range between a maximum threshold and a minimum threshold, wherein the first range is based upon values of historic readings of the sensor; and determining whether the readings included in the first dataset comprise values within the first range; wherein the first check comprises an averaging check that includes the steps of: calculate average values for the readings in the first dataset, wherein the average value comprises the averaging of corresponding readings from the sensor and at least one other sensor of the same type; defining a second range about the calculated average values in which a positive offset from the calculated average values comprises a maximum threshold, and a negative offset from the calculated average values comprises a minimum threshold; and determining whether the readings included in the first dataset comprise values within the second range; and wherein the first check comprises a data check that includes determining whether a sequential plot of the readings of the first dataset over the first lookback period comprises a profile indicative of a data irregularity that includes at least one of a drift, a shift, and a spike.

17 . The method according to claim 7 , wherein the simulation of the operation of the thermal generating unit comprises a tuned model of the thermal generating unit;
further comprising the steps of:
sensing and collecting measured values for a plurality of the operating parameters of the thermal generating unit; and
tuning a model of the thermal generating unit so to configure the tuned model of the thermal generating asset, wherein the tuning comprises a data reconciliation process wherein the measured values for selected ones of the operating parameters are compared to predicted values for the selected ones of the operating parameter so to determine a differential therebetween upon which the tuning of the model is based.

18 . The method according to claim 17 , wherein the model of the thermal generating unit comprises a physics-based model, an

Assignees

• Gen Electric

Inventors

• Wichmann Lisa Anne
• Pandey Achalesh Kumar
• Raczynski Christopher Michael

Classifications

• G05B13/042 Primary

  in which a parameter or coefficient is automatically adjusted to optimise the performance · CPC title

• G05F1/66

  Regulating electric power · CPC title