Method for estimating oil/gas production using statistical learning models

1. A method for allocating production for an oil or gas well, comprising:

receiving input data corresponding to a well on a lease, wherein the input data comprises well test data, completion/recompletion data, and lease production data;

detecting, through an outlier detection analysis, one or more outlying data points in the well test data;

detecting, through a change point detection analysis of the well test data, one or more data points indicative of an intervention, wherein each data point corresponds to a distinct time period;

generating a decline curve for each time period in between the one or more data points indicative of an intervention;

following generation of the decline curve, disregarding the one or more outlying data points;

determining the production of the well for a predetermined period of time by integrating the decline curve over the predetermined period of time;

calculating the allocated production for the well by multiplying the ratio of the production of the well to the sum of the production for all wells in the lease by a production per lease value;

training a machine learning algorithm to predict decline curves based on at least a portion of the input data; and

predicting, using the machine learning algorithm, a decline curve for one or more wells lacking well test data but having completion data.

2. The method as recited in claim 1 , further comprising:

generating completion data for the one or more wells lacking well test data using one or more statistical methods.

3. A method for allocating production for an oil or gas well, comprising:

receiving input data corresponding to a well on a lease, wherein the input data comprises well test data, completion/recompletion data, and lease production data;

detecting, through an outlier detection analysis, one or more outlying data points in the well test data;

detecting, through a change point detection analysis of the well test data, one or more data points indicative of an intervention, wherein each data point corresponds to a distinct time period;

generating a decline curve for each time period in between the one or more data points indicative of an intervention;

following generation of the decline curve, disregarding the one or more outlying data points;

determining the production of the well for a predetermined period of time by integrating the decline curve over the predetermined period of time;

calculating the allocated production for the well by multiplying the ratio of the production of the well to the sum of the production for all wells in the lease by a production per lease value; and

calculating cumulative sum of differences between the allocated production for the well and average production for all wells in the lease;

detecting local minima of the cumulative sum of differences; and

designating those local minima as the one or more data points indicative of an intervention.

4. A computer program product for allocating production for an oil or gas well, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to:

receive input data corresponding to a well on a lease, wherein the input data comprises well test data, completion/recompletion data, and lease production data;

detect, through an outlier detection analysis, one or more outlying data points in the well test data;

detect, through a change point detection analysis of the well test data, one or more data points indicative of an intervention, wherein each data point corresponds to a distinct time period;

generate a decline curve for each time period in between the one or more data points indicative of an intervention;

following generation of the decline curve, disregarding the one or more outlying data points;

determine the production of the well for a predetermined period of time by integrating the decline curve over the predetermined period of time; and

calculate the allocated production for the well by multiplying the ratio of the production of the well to the sum of the production for all wells in the lease by a production per lease value.

5. The computer program product as recited in claim 4 , wherein the processor is further caused to:

training a machine learning algorithm to predict decline curves based on at least a portion of the input data; and

predicting, using the machine learning algorithm, a decline curve for one or more wells lacking well test data but having completion data.

6. The computer program product as recited in claim 5 , wherein the processor is further caused to:

generate estimated completion data using one or more statistical methods; and input the estimated completion data into the decline curve.

7. The computer program product as recited in claim 4 , wherein the outlier detection analysis is a “leave-one-out” design strategy.

8. The computer program product as recited in claim 4 , wherein the change point detection analysis further causes the processor to:

calculating cumulative sum of differences between the measured production values and average production for all wells in the lease;

detect local minima of the cumulative sum of differences; and

designate those local minima as the one or more data points indicative of an intervention.

9. The computer program product as recited in claim 4 , wherein the generated decline curve is exponential.

10. The computer program product as recited in claim 4 , wherein the generated decline curve is hyperbolic.

11. The computer program product as recited in claim 4 , wherein the generated decline curve comprises one or more curve segments, wherein each curve segment comprises unique decline parameters.