Application of marsh funnel through use of trained algorithm
Inventors
Ofoche, Paul • Noynaert, Samuel F.
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12399097-B2
Publication Date
2025-08-26
Expiration Date
Abstract
A method includes obtaining a density of a fluid and obtaining a Marsh funnel time associated with the fluid. The density of the fluid and the Marsh funnel time is provided to a processor. The processor derives properties of the fluid from the fluid density and the Marsh funnel time. A machine-learning algorithm is applied to the properties of the fluid. The machine-learning algorithm determines a plastic viscosity and a yield point of the fluid. Output of the machine-learning algorithm is stored for future use. Properties of the drilling fluid are adjusted based on the output of the machine learning algorithm.
Core Innovation
The invention determines at least one property of a fluid using a Marsh funnel time. The Marsh funnel time is provided to a processor coupled to a memory and a data acquisition unit, and a machine-learning algorithm is applied to determine the at least one property, wherein the at least one property comprises a synthetic set of dial readings for the fluid.
The output of the machine-learning algorithm is stored for future use and the at least one property of the fluid is adjusted based on the output. In monitoring embodiments, an input comprising a Marsh funnel time associated with the fluid is received by a data acquisition system comprising a processor and a memory, and a synthetic set of dial readings is calculated based on a trained model using the Marsh funnel time.
The invention supports near real time adjustment and monitoring of fluid properties using the synthetic set of dial readings. A system includes a Marsh funnel configured to receive the fluid, a data-acquisition unit coupled to the Marsh funnel to collect the Marsh funnel time, and a processor coupled to the data-acquisition unit and to a drilling fluid system to adjust the at least one fluid parameter based on the synthetic set of dial readings.
Claims Coverage
The document includes three independent claims that broadly cover a method, a system, and a monitoring method using Marsh funnel time and synthetic set of dial readings.
Marsh funnel time to synthetic dial readings mapping via machine learning
Obtaining a Marsh funnel time associated with the fluid and applying, via a processor, a machine-learning algorithm to the Marsh funnel time to determine at least one property, wherein the at least one property comprises a synthetic set of dial readings for the fluid.
Storing machine-learning output for future use in fluid-property adjustment
Storing output of the machine-learning algorithm for future use and adjusting, via the processor, the at least one property of the fluid based on the output of the machine learning algorithm.
Near real time monitoring using a trained model to calculate synthetic dial readings
Receiving, by a data acquisition system comprising a processor and a memory, an input comprising a Marsh funnel time associated with the fluid; calculating, by the processor, a synthetic set of dial readings based, at least in part, on a trained model of the data acquisition system using the Marsh funnel time; and adjusting, by the processor, in near real time the at least one fluid property of the fluid based at least in part upon the synthetic set of dial readings.
System coupling Marsh funnel measurement to drilling-fluid parameter adjustment
Providing a system comprising a Marsh funnel configured to receive the fluid; a data-acquisition unit operatively coupled to the Marsh funnel and configured to collect a Marsh funnel time of the fluid; and a processor comprising memory and coupled to the data-acquisition unit, the processor configured to determine the at least one fluid parameter using a machine learning algorithm applied to the fluid Marsh funnel time, wherein the at least one fluid parameter comprises a synthetic set of dial readings, and wherein the processor is operatively coupled to a drilling fluid system to adjust the at least one fluid parameter based at least in part upon the synthetic set of dial readings.
Across the independent claims, the core coverage is the use of a machine-learning algorithm or trained model to convert a Marsh funnel time into a synthetic set of dial readings, followed by storing and/or using that output to adjust fluid properties, with a system embodiment coupled to a drilling fluid system and a monitoring embodiment emphasizing near real time adjustment.
Stated Advantages
Continuous/non-intrusive monitoring is described in the partial content as an advantage.
Site decision support is described in the partial content as an advantage.
Predictive accuracy comparisons to conventional rheometers are described in the partial content.
Documented Applications
Monitoring and adjusting drilling-fluid properties using Marsh funnel time, synthetic dial readings, and a processor coupled to a drilling mud system.
Interested in licensing this patent?