Noise detection and suppression system and method for wellbore telemetry

• US 6,023,658 A
• Filed: 04/09/1997
• Issued: 02/08/2000
• Est. Priority Date: 04/09/1996
• Status: Expired due to Fees

• Current Assignee: Schlumberger Technology Corporation (SLB)
• Original Assignee: Schlumberger Technology Corporation (SLB)
• Inventors: Jeffryes, Benjamin P.
• Primary Examiner(s): Oda, Christine K.
• Assistant Examiner(s): Taylor, Victor J.
• Application Number:
  US08/831,548
• Time in Patent Office: 1,035 Days
• Field of Search: 367/81, 367/82, 367/83, 191/12, 702/16, 702/6, 702/10
• US Class Current: 702/16
• CPC Class Codes: E21B 47/18 through the well fluid , e.g. mud pressure pulse telemetry
• CPC Class Codes: G01V 1/364 Seismic filtering G01V1/37 takes precedence
• CPC Class Codes: G01V 2210/21 Frequency-domain filtering, e.g. band pass
• CPC Class Codes: G01V 2210/324 Filtering

Abstract

Methods and apparatus are described for identifying and removing noise from borehole telemetry signals. A spectrum of the signal and of an approximated noise-free spectrum of the signals are generated and compared so as to identify frequencies at which noise occurs. The approximated spectrum is preferably generated by an averaging or integrating process which essentially results in a smoothened variant of the signal spectrum. The comparison includes dividing the signal spectrum by the approximated spectrum, and setting a threshold in the resulting dimensionless peak spectrum. The remaining parts of the peak spectrum are identified as noise. The dimensionless spectrum is found to have major advantages when having to identify noise frequencies.

Claims

1. A noise filtering apparatus for wellbore telemetry data signals, comprising:

a receiver located on the surface and adapted to receive a wellbore telemetry data signal, said telemetry data signal including data which represents one or more downhole measurements conducted downhole during a drilling operation, the data being unknown to the receiver prior to reception;

a transformer in communication with said receiver and adapted to transform at least part of the telemetry data signal to generate therefrom a spectrum of at least part of the telemetry data signal;

an approximator in communication with said transformer adapted to generate an approximation of a noise-free spectrum of the at least part of the telemetry data signal;

a ratio calculator in communication with said approximator adapted to generate a dimensionless spectrum by calculating the ratio of the spectrum of at least part of the telemetry data signal and the approximation of a noise-free spectrum; and

an identifier in communication with said ratio calculator adapted to analyze the dimensionless spectrum and select frequencies from said dimensionless spectrum that are characterized by a substantial level of noise.

2. The apparatus in accordance with claim 1, wherein the approximator comprises a weighted means based filter, and the weighted means based filter is applied to the spectrum of at least part of the telemetry data signal to produce therefrom the approximation of a noise-free spectrum of the at least part of the telemetry data signal.

3. The apparatus in accordance with claim 1, wherein the approximator comprises a median filter, and the median filter is applied to the spectrum of at least part of the telemetry data signal to produce therefrom the approximation of a noise-free spectrum of the at least part of the telemetry data signal.

4. The apparatus in accordance with claim 1, further comprising means for filtering the selected frequencies from the telemetry data signal.

5. The apparatus in accordance with claim 1, further comprising means for removing spike-type noise from the telemetry data signal prior to the generation of the dimensionless spectrum.

6. The apparatus in accordance with claim 1, further comprising means for setting a predetermined threshold in the dimensionless spectrum, and means for discarding all values of the dimensionless spectrum which have a value below the predetermined threshold.

20. The apparatus of claim 1, wherein the transformer, approximator, ratio calculator, and identifier reside in a general purpose computing machine, and communication between the transformer, approximator, ratio calculator, and identifier is carried out in said general purpose computing machine.

21. The apparatus of claim 1, further comprising:

an electromechanical transducer adapted to detect a raw signal and to convert the raw signal into an analogue electrical signal, the raw signal consisting of acoustic energy transmitted through a drilling fluid; and

a processor in communication with said electro mechanical transducer and said receiver, said processor adapted and configured to sample and process the analogue signal to produce the telemetry data signal, and to send the telemetry data signal to said receiver.

7. A wellbore telemetry apparatus for gathering data related to subsurface conditions and for transmitting said data to the surface, said apparatus comprising:

a first transducer located in a well adapted and configured to convert data gathered in the well into signals to be in operation transmitted to said surface;

a second transducer located on said surface adapted and configured to reconvert said signals into processable data; and

a processor in communication with said second transducer adapted and configured to process said processable data, the processor characterized in that said processor comprises a transformer for generating a signal spectrum of at least a part of the processable data, an approximator for generating an approximation of a noise-free spectrum of said at least a part of said processable data, and a ratio generator adapted to compute the ratio of said signal spectrum and said approximation of a noise-free spectrum so as to identify noise frequencies.

8. A method for identifying a frequency or range of frequencies of noise in a wellbore telemetry data signal, comprising the steps of:

receiving a wellbore telemetry data signal using a receiver located on the surface, said telemetry data signal including data which represents one or more downhole measurements gathered downhole during a drilling operation, the data being unknown to the receiver on the surface prior to reception;

transforming at least part of the telemetry data signal to generate therefrom a spectrum of at least part of the telemetry data signal;

generating an approximation of a noise-free spectrum of the at least part of the telemetry data signal;

generating a dimensionless spectrum by calculating a ratio of the spectrum of at least part of the telemetry data signal and the approximation of a noise-free spectrum; and

identifying noise frequencies by analyzing the dimensionless spectrum and selecting frequencies from said dimensionless spectrum.

9. The method of claim 8, wherein the step of generating an approximation of a noise-free spectrum comprises applying a weighted means based filter to the spectrum of at least part of the telemetry data signal to produce therefrom the approximation of a noise-free spectrum of the at least part of the telemetry data signal.

10. The method of claim 8, wherein the step of generating an approximation of a noise-free spectrum comprises applying a weighted median based filter to the spectrum of at least part of the telemetry data signal to produce therefrom the approximation of a noise-free spectrum of the at least part of the telemetry data signal.

11. The method of claim 8, wherein the step of identifying the noise frequencies includes the step discarding all values of the dimensionless spectrum which have a value below a predetermined threshold.

12. The method of claim 8, wherein the step of transforming at least part of the telemetry data signal comprises performing a Fourier transformation or a mathematical equivalent thereof.

13. The method of claim 8, wherein prior to the step of transforming at least part of the telemetry data signal, trends in said telemetry data signal are removed.

14. The method of claim 8, wherein prior to the step of transforming at least part of the telemetry data signal, spikes in said telemetry data signal are removed.

15. The method of claim 8, wherein prior to the step of transforming at least part of the telemetry data signal, a zero signal is added to both sides of the at least a part of the telemetry data signal.

16. The method of claim 8, wherein prior to the step of transforming at least part of the telemetry data signal, the at least part of the telemetry data signal is multiplied by a window function with smooth edges.

17. The method of claim 8, further comprising the step of removing at least some noise from the spectrum of at least part of the telemetry data signal by suppressing noise frequencies as identified in the step of identifying noise frequencies.

18. The method of claim 17, wherein the noise frequencies are suppressed by using a noise matrix N derived from two linear independent state vectors for each identified noise frequency, and minimizing the residual between the gathered signal and a noise vector derived from multiplying said noise matrix N with a vector of unknown noise amplitudes.

19. The method of claim 8, including the step of transmitting control signals for a downhole apparatus.

22. The method of claim 8, further comprising the steps of:

detecting a raw signal using an electro-mechanical transducer, the raw signal consisting of acoustic energy transmitted through a drilling fluid;

converting the raw signal into an analogue electrical signal;

sampling and processing the analogue signal to produce the telemetry data signal; and

sending the telemetry data signal to said receiver.