Systems, methods, and computer medium to provide entropy based characterization of multiphase flow
Systems, computer-implemented methods, and non-transitory computer-readable medium having a stored computer program provide characterization of multiphase fluid flow (MPF) using approximate entropy calculation techniques to enhance measuring and monitoring of a flow regime in a segment of pipe for hydrocarbon-production operations. The systems and methods can be optimized using principal component analysis.
1. A multiphase fluid flow (MPF) characterization system to enhance measuring and monitoring of a flow regime in a segment of pipe for hydrocarbon-production operations, the system comprising:
an acoustic emission sensor disposed proximate to the segment of pipe and operable to receive an acoustic emission from a MPF, the segment of pipe operable to support the MPF in hydrocarbon-production operations including at least two physical phases, and the acoustic emission sensor operable to convert the received acoustic emission to an electrical signal;
a processing unit, including a processor, operable to receive the electrical signal and transform the electrical signal to characterize the MPF, the processing unit in communication with and comprising:
non-transitory, tangible memory medium in communication with the processor having a set of stored instructions, the set of stored instructions being executable by the processor and including the steps of:
segmenting the electrical signal into short term, medium term, and long term time series;
assigning positive real numbers to the time series, the positive real numbers including larger values and smaller values, the larger values corresponding to process randomness, and the smaller values corresponding to instances of recognizable patterns in the electrical signal;
categorizing certain positive real numbers as outlier values;
calculating short term, medium term, and long term approximate entropy values for the MPF responsive to the short term, medium term, and long term time series from the electrical signal by performing the following steps:
assigning each of the short term, medium term, and long term time series of data to variables y( 1 ), y( 2 ), . . . , y(N) where N raw data values are measured at equally spaced time,
fixing an integer m, and a positive real number r, where m represents the length of compared run of data, and where r specifies filtering level,
forming a sequence of vectors x( 1 ), x( 2 ), . . . , (N−m +1) in R m , where real m-dimensional space defined by x(i)=[y(i), y(i+1), . . . , y(i+m−1)]
using the sequence x( 1 ), x( 2 ), . . . , x(N−m+1) to construct for each i, where 1≤i≤N−m+1, and
C i m (r)=(number of x(j)such that d[x(i), x(j)]<r)/(N−m+1 where d[x, x′]=max a |y(a)|, and the u (a) are the m scalar components of x, and d represents the distance between the vectors x(i) and x(j), given by the maximum difference in their respective scalar components,
calculating
Φ
m
(
r
)
=
(
N
-
m
+
1
)
-
1
∑
i
=
1
N
-
m
+
1
log
(
C
i
m
(
r
)
)
calculating each tort term, medium term, and long term time. series approximate entropy as,
ApEn=Φ m ( r )=Φ m+1 ( r );
comparing the short term, medium term, and long term approximate entropy values for the MPF to pre-determined short term, medium term, and long term approximate entropy values; and
determining characteristics of the MPF responsive to similarities between the short term, medium term, and long term approximate entropy values for the MPF and the pre-determined short term, medium term, and long term approximate entropy values; and
a user interface coupled to the processing unit, the user interface operable to accept user inputs to control the processing unit, and operable to display the characteristics of the MPF to a user.
2. The system of claim 1 further comprising a database with pre-determined short term, medium term, and long term approximate entropy values for a variety of MPF flow regimes.
3. The system of claim 1 further comprising a preamplifier coupled to the acoustic emission sensor, and operable to receive and amplify the electrical signal from the acoustic emission sensor.
4. The system of claim 3 further comprising a band-pass signal filter coupled to the preamplifier,
the band-pass signal filter operable to receive an amplified electrical signal from the preamplifier, and
further operable to remove acoustic background noise from useful MPF acoustic information contained within the amplified electrical signal, responsive to programmed cutoff frequencies in the band-pass signal filter derived from an operating frequency and bandwidth of the acoustic emission sensor.
5. The system of claim 4 further comprising an analog-to-digital converter coupled to the band-pass signal filter, the analog-to-digital converter operable to receive from the band-pass signal filter the useful MPF acoustic information, and operable to convert the useful MPF acoustic information to a digital signal.
6. The system of claim 1 , where the processing unit further is operable to execute a set of instructions to conduct a principal component analysis on the system including the steps of:
gathering acoustic emission data under a variety of flow parameters in situations in which an appropriate Reynolds number is known for the MPF for which data is being gathered;
forming time series of acoustic waveforms;
performing a Fourier Transformation on the data, the data being converted into measurements of acoustic power as a function of frequency;
executing a suite of measurements using a test matrix including different conditions of the MPF including at least one variable selected from the group consisting of: stepped values of watercut, stepped values of total liquid flow, and MPF regimes; and
post-processing the data by applying principal component analysis to the data to determine measurable frequencies relevant to determining the characteristics of the MPF.
7. The system of claim 6 further comprising an optimized acoustic emission sensor, where the optimized acoustic emission sensor is operable to receive the frequencies determined by the principal component analysis to be relevant to determining the characteristics of the MPF.
8. The system of claim 1 , where the acoustic emission sensor comprises a first acoustic emission sensor, and where the system further comprises a second acoustic emission sensor disposed proximate to the segment of pipe and operable to receive an acoustic emission from the MPF, the second acoustic emission sensor further operable to convert the received acoustic emission to an electrical signal.
9. The system of claim 8 , where the second acoustic emission sensor is disposed at a distance D from the first acoustic emission sensor,
where the distance D is operable to allow coherent measurements of the MPF in a substantially similar state at both the first acoustic emission sensor and the second acoustic emission sensor, and
where an accurate measurement of flow velocity of the MPF is obtained by dividing the distance D by a difference in time between a first time at which the MPF passes the first acoustic emission sensor and a second time at which the MPF passes the second acoustic emission sensor.
10. The system of claim 8 , where the processing unit further is operable to execute a set of instructions to conduct a principal component analysis on the system including the steps of:
gathering acoustic emission data under a variety of flow parameters in situations in which an appropriate Reynolds number is known for the MPF for which data is being gathered;
forming time series of acoustic waveforms;
performing a Fourier Transformation on the data, the data being converted into measurements of acoustic power as a function of frequency;
executing a suite of measurements using a test matrix including different conditions of the MPF including at least one variable selected from the group consisting of: stepped values of watercut, stepped values of total liquid flow, and MPF regimes; and
post-processing the data by applying principal component analysis to the data to determine measurable frequencies relevant to determining the characteristics of the MPF.
11. The system of claim 10 further comprising an optimized acoustic emission sensor, where the optimized acoustic emission sensor is operable to receive the frequencies determined by the principal component analysis to be relevant to determining the characteristics of the MPF.
12. A method for characterizing multiphase fluid flow (MPF) to enhance measuring and monitoring of a flow regime in a segment of pipe for hydrocarbon-production operations, the method comprising the steps of:
sensing an acoustic emission from a MPF, the segment of pipe operable to support the MPF in hydrocarbon-production operations including at least two physical phases;
converting the acoustic emission to an electrical signal; and
a multi-phase flow system comprising a processing unit, including a processor, operable to receive the electrical signal and transform the electrical signal to characterize the MPF, the processing unit in communication with and comprising:
non-transitory, tangible memory medium in communication with the processor having a set of stored instructions, the set of stored instructions being executable by the processor and including the steps of:
segmenting the electrical signal into short term, medium term, and long term time series;
assigning positive real numbers to the time series, the positive real numbers including larger values and smaller values, the larger values corresponding to process randomness, and the smaller values corresponding to instances of recognizable patterns in the electrical signal;
categorizing certain positive real numbers as outlier values;
calculating short term, medium term, and long term approximate entropy values for the MPF responsive to the short term, medium term, and long term time series from the electrical signal by performing the following steps:
assigning each of the short term, medium term, and long term time series of data to variables y( 1 ), y( 2 ), . . . , y(N) where N raw data values are measured at equally spaced time,
fixing an integer m, and a positive real number r, where m represents the length of compared run of data, and where r specifies filtering level,
forming a sequence of vectors x( 1 ), x( 2 ), . . . , x(N−m+1) in R m , where real m-dimensional space defined by x(i)=[y(i), y(i+1), . . . , y(i+m+1)]
using the sequence x( 1 ), x( 2 ), . . . , x(N−m+1) to construct for each i, where 1≤i≤N−m+1, and
C i m (r) =(number of x(j) such that d[x(i), x(j)]<r)/(N−m+1 where d[x, x*]=max a |y(a)−y*(a)|, and the u(a) are the m scalar components of x, and d represents the distance between the vectors x(i) and x(j), given by the maximum difference in their respective scalar components,
calculating,
Φ
m
(
r
)
=
(
N
-
m
+
1
)
-
1
∑
i
=
1
N
-
m
+
1
log
(
C
i
m
(
r
)
)
calculating each of the short term, medium term, and long term time series approximate entropy as
ApEn=Φ m ( r )−Φ m+1 ( r );
comparing the short term, medium term, and long term approximate entropy values for the MPF to pre-determined short term, medium term, and long term approximate entropy values; and
determining characteristics of the MPF responsive to similarities between the short term, medium term, and long term approximate entropy values for the MPF and the pre-determined short term, medium term, and long term approximate entropy values.
13. The method of claim 12 , further comprising the step of displaying the characteristics of the MPF on a user interface, where the user interface is operable to graphically represent at least one flow regime.
14. The method of claim 12 , further comprising the step of preamplifying the electrical signal before the step of segmenting the electrical signal.
15. The method of claim 14 further comprising the step of filtering the electrical signal, before segmenting the electrical signal, responsive to programmed cutoff frequencies in a band-pass signal filter derived from an operating frequency and bandwidth of an acoustic emission sensor.
16. The method of claim 15 further comprising the step of converting the electrical signal to a digital signal, before segmenting the electrical signal.
17. The method of claim 12 , further comprising the step of conducting a principal component analysis, where the principal component analysis comprises the steps of:
gathering acoustic emission data under a variety of flow parameters in situations in which an appropriate Reynolds number is known for the MPF for which data is being gathered;
forming time series of acoustic waveforms;
performing a Fourier Transformation on the data, the data being converted into measurements of acoustic power as a function of frequency;
executing a suite of measurements using a test matrix including different conditions of the MPF including at least one variable selected from the group consisting of: stepped values of watercut, stepped values of total liquid flow, and multiphase flow patterns; and
post-processing the data by applying principal component analysis to the data to determine measurable frequencies relevant to determining the characteristics of the MPF.
18. The method of claim 17 further comprising the step of optimizing the step of sensing an acoustic emission from the MPF to receive the frequencies determined by the principal component analysis to be relevant to determining the characteristics of the MPF.
19. The method of claim 12 , where the step of sensing an acoustic emission comprises the step of sensing a first acoustic emission, and further comprises the step of sensing a second acoustic emission from the MPF, the second acoustic emission being sensed simultaneously with and at a distance D from the first acoustic emission.
20. The method of claim 19 , further comprising the step of calculating an accurate measurement of flow velocity of the MPF in response to the distance D and sensing the first acoustic emission and sensing the second acoustic emission.
21. The method of claim 19 , further comprising the step of conducting a principal component analysis, where the principal component analysis comprises the steps of:
gathering acoustic emission data under a variety of flow parameters in situations in which an appropriate Reynolds number is known for the MPF for which data is being gathered;
forming time series of acoustic waveforms;
performing a Fourier Transformation on the data, the data being converted into measurements of acoustic power as a function of frequency;
executing a suite of measurements using a test matrix including different conditions of the MPF including at least one variable selected from the group consisting of: stepped values of watercut, stepped values of total liquid flow, and multiphase flow patterns; and
post-processing the data by applying principal component analysis to the data to determine measurable frequencies relevant to determining the characteristics of the MPF.
22. The method of claim 21 further comprising the step of optimizing the step of sensing an acoustic emission from the MPF to receive the frequencies determined by the principal component analysis to be relevant to determining the characteristics of the MPF.