Predicting clean fluid composition and properties with a rapid formation tester pumpout
A method and system for estimating clean fluid composition and properties. A method may comprise disposing a downhole fluid sampling tool into a wellbore, wherein the downhole fluid sampling tool comprises optical instrumentation, obtaining a fluid sample with the downhole fluid sampling tool, wherein the fluid sample comprises a reservoir fluid contaminated with a well fluid, identifying input parameters from at least one sensor response on the optical instrumentation, and predicting a clean fluid sample of the reservoir fluid using an asymptote of a dimensional reduction analysis and equation of state. A system may comprise a downhole fluid sampling tool and a processing unit. The downhole fluid sampling tool may further comprise an optical instrumentation operable to obtain fluid samples of a reservoir fluid contaminated with a well fluid while the downhole fluid sampling tool is disposed in a wellbore.
1. A method for estimating clean fluid composition and properties, comprising:
disposing a downhole fluid sampling tool into a wellbore, wherein the downhole fluid sampling tool comprises a fluid analysis module;
obtaining a fluid sample that has a contamination level for which a fluid property is not representative of a pure formation fluid with the downhole fluid sampling tool;
identifying input parameters from at least one sensor response on the fluid analysis module, wherein the input parameters comprise fluid properties indicative of the fluid sample and mud filtrate composition;
applying a robust regression method on at least one sensor response, wherein the robust regression method is computed as scores fit =A+BXv β , where A and B are constants determined by a regression method, v is a formation fluid pumpout accumulated volume, and β is a density decay rate; and
predicting a clean fluid sample of the fluid sample using a dimensional reduction analysis and an equation of state, wherein the equation of state utilizes one or more component mole fractions of the fluid sample.
2. The method of claim 1 , wherein the contamination level is not less than five percent of the fluid sample.
3. The method of claim 1 , wherein the fluid analysis module measures gas-to-oil ratio, fluid density, fluid viscosity, formation pressure, or formation temperature.
4. The method of claim 1 , further comprising despiking the at least one sensor response.
5. The method of claim 1 , wherein the contamination level of a mud filtrate is twenty percent or higher of the fluid sample.
6. The method of claim 1 , further comprising transforming the at least one sensor response by a principal component analysis.
7. The method of claim 1 , further comprising performing a robust principal component analysis on the at least one sensor response.
8. The method of claim 1 , wherein obtaining the fluid sample comprising operating the downhole fluid sampling tool in the wellbore to obtain the fluid sample.
9. A system for estimating clean fluid composition, comprising:
a downhole fluid sampling tool comprising a fluid analysis module that obtains fluid samples that have a contamination level that is not less than five percent of the fluid sample while the downhole fluid sampling tool is disposed in a wellbore; and
a processing unit configured to:
predict a clean fluid sample of the fluid from at least one sensor response on the fluid analysis module using an asymptote analysis and an equation of state, wherein the equation of state utilizes one or more component mole fractions of the fluid sample; and
apply a robust regression method on at least one sensor response, wherein the robust regression method is computed as scores fit =A+BXv β , where A and B are constants determined by a regression method, v is a formation fluid pumpout accumulated volume, and β is a density decay rate.
10. The system of claim 9 , wherein the fluid analysis module measures gas-to-oil ratio, fluid density, fluid viscosity, formation pressure, or formation temperature.
11. The system of claim 9 , wherein the processing unit is further configured to despike the at least one sensor response.
12. The system of claim 9 , wherein the contamination level of a mud filtrate is twenty percent or higher of the fluid sample.
13. The system of claim 9 , wherein the processing unit is further configured to transform the at least one sensor response to a principal component analysis space.
14. The system of claim 9 , wherein the processing unit is further configured to perform a robust principal component analysis on the at least one sensor response.
15. A system for estimating clean fluid composition, comprising:
a downhole fluid sampling tool comprising a fluid analysis module that obtains a fluid sample while the downhole fluid sampling tool is disposed in a wellbore; and
a processing unit configured to:
predict a clean fluid sample of the fluid sample from at least one sensor response on the fluid analysis module using an asymptote analysis and an equation of state, wherein the equation of state utilizes one or more component mole fractions of the fluid sample, and
apply a robust regression method on at least one sensor response, wherein the robust regression method is computed as scores fit =A+BXv β , where A and B are constants determined by a regression method, v is a formation fluid pumpout accumulated volume, and β is a density decay rate.
16. The system of claim 15 , wherein the fluid analysis module measures gas-to-oil ratio, fluid density, fluid viscosity, formation pressure, or formation temperature.
17. The system of claim 15 , wherein the processing unit is further configured to despike the at least one sensor response.
18. The system of claim 17 , wherein the contamination level of a mud filtrate is twenty percent or higher of the fluid sample.
19. The system of claim 15 , wherein the fluid sample has a contamination level that not less than five percent of the fluid sample.
20. The system of claim 15 , wherein the processing unit is further configured to transform the at least one sensor response to a principal component analysis space.