Determination of asphaltene onset condition of reservoir fluids during downhole fluid analysis

A system comprising:

• a downhole acquisition tool comprising a downhole fluid analysis module configured to obtain downhole fluid analysis data corresponding to a reservoir fluid;
• and
• processing circuitry configured to:
• receive first historical downhole fluid analysis data and first historical laboratory measurements, wherein the first historical downhole fluid analysis data comprises a pressure, a reservoir temperature, a carbon content, a fluid density, a fluid viscosity, or any combination thereof;
• identify a portion of the first historical downhole fluid analysis data by removing one or more first datasets outside a range of values and removing one or more second datasets that are not part of the first historical laboratory measurements;
• partition the portion of the first historical downhole fluid analysis data into training data and testing data, wherein the testing data comprises at least one dataset corresponding to a first historical laboratory measurement of the first historical laboratory measurements classified as having asphaltene;
• identify the downhole fluid analysis data from the testing data;
• implement a first statistical model trained using the training data to identify, based on the downhole fluid analysis data corresponding to the reservoir fluid, a likelihood of whether the reservoir fluid contains problematic asphaltenes or an asphaltene onset pressure (AOP) of the reservoir fluid;
• and
• verify the first statistical model is accurate based on the testing data.

A method comprising:

• collecting, for respective samples of a plurality of samples of reservoir fluids, historical downhole fluid analysis data and historical laboratory measurements, wherein the historical downhole fluid analysis data comprises a pressure, a reservoir temperature, a carbon content, a fluid density, a fluid viscosity, or any combination thereof;
• identifying a portion of the historical downhole fluid analysis data by removing one or more first datasets outside a range of values and removing one or more second datasets that are not part of the historical laboratory measurements;
• partitioning the portion of the historical downhole fluid analysis data into training data and testing data, wherein the testing data comprises at least one dataset corresponding to a first historical laboratory measurement of the historical laboratory measurements classified as having asphaltene and at least one dataset corresponding to a second historical laboratory measurement of the historical laboratory measurements classified as not having asphaltene;
• identifying downhole fluid analysis data from the testing data, wherein the downhole fluid analysis data corresponds to a reservoir fluid that was obtained by a downhole acquisition tool positioned in a wellbore;
• training a model using the training data to identify a likelihood of whether a new reservoir fluid contains problematic asphaltenes or an asphaltene onset pressure (AOP) of the new reservoir fluid;
• and
• verifying the model is accurate based on the testing data by comparing a result of the model to the at least one dataset corresponding to the first historical laboratory measurement of the historical laboratory measurements classified as having asphaltene and to the at least one dataset corresponding to the second historical laboratory measurement of the historical laboratory measurements classified as not having asphaltene.