Abstract: A method of controlling a pumping stage of a fracturing fleet at a wellsite with a set of diesel pumps and at least one electric pump to smooth the flowrate transition to an operational setpoint with a higher flowrate. An optimization process communicatively connected to the plurality of pumping units can iterate an interim setpoint with a transfer function model to each pumping unit. The transfer function model can generate a smooth flowrate transition by decreasing the flowrate of at least one pump unit while increasing the flowrate to the remaining pump units.

Abstract: A method of controlling a pumping stage of a fracturing fleet at a wellsite with at least two electric frac pumps to increase the efficiency of the electric frac pumps comprises communicating an interim flowrate to the electric frac pumps and calculating an efficiency value from measured power from the VFD and measured flow rate, intake pressure, and discharge pressure. The method can increase a total efficiency of the fracturing fleet above a threshold efficiency value by iterating the interim setpoint from a first interim setpoint to a second interim setpoint for the electric frac pumps, wherein the second interim setpoint decreases the flowrate to the less efficient electric frac pumps and increases the flowrate an equal amount to the more efficient electric frac pumps.

Abstract: Aspects of the subject technology relate to systems and methods for optimizing multi-unit pumping operations at a well site. Systems and methods are provided for receiving sensor data from a hydraulic fracturing fleet equipment at an equipment system, designating an event as being flagged based on the sensor data from the hydraulic fracturing fleet equipment, determining a physical action based on the flagged event and a priority list of actions, and providing instructions to a first pump of the hydraulic fracturing fleet equipment to perform the physical action based on the flagged event and the priority list of actions.

Abstract: A system is provided including at least one pump for pumping a fluid into a wellbore, a pressure sensor provided at a wellhead of the wellbore for measuring a backpressure of the fluid being pumped into the wellbore, and a diagnostic manager. The diagnostic manager obtains pressure data associated with a pressure signal from the pressure sensor, wherein the pressure data includes pressure measurements of the fluid over a selected time period. The diagnostic manager converts, based on the pressure data, at least a portion of the pressure signal into frequency domain. The diagnostic manager detects a change in frequency of the pressure signal in the Fourier spectrum and determines that a fault associated with the wellbore has occurred based on the changed frequency of the pressure signal.

Abstract: A machine learning model trained to predict offset pressure (“predictor”) is used to generate a pumping schedule that mitigates well interference during a hydraulic fracturing treatment operation. Diverse candidate pumping schedules are generated according to pumping schedule constraints. Feature inputs are populated with the candidate pumping schedules and with static and dynamic features related to the operation. The feature inputs are fed into one or more instances of the predictor to obtain offset pressure predictions at a prediction horizon for which the predictor was configured. The offset pressure predictions are evaluated to identify the one that best satisfies a well interference mitigation objective and the corresponding candidate pumping schedule is identified for well interference mitigation.

Abstract: A system is provided including at least one pump for pumping a fluid into a wellbore, a pressure sensor provided at a wellhead of the wellbore for measuring a backpressure of the fluid being pumped into the wellbore, and a diagnostic manager. The diagnostic manager obtains pressure data associated with a pressure signal from the pressure sensor, wherein the pressure data includes pressure measurements of the fluid over a selected time period. The diagnostic manager converts, based on the pressure data, at least a portion of the pressure signal into frequency domain. The diagnostic manager detects a change in frequency of the pressure signal in the Fourier spectrum and determines that a fault associated with the wellbore has occurred based on the changed frequency of the pressure signal.

Abstract: Aspects of the subject technology relate to systems and methods for optimizing multi-unit pumping operations at a well site. Systems and methods are provided for receiving sensor data from a hydraulic fracturing fleet equipment at an equipment system, designating an event as being flagged based on the sensor data from the hydraulic fracturing fleet equipment, determining a physical action based on the flagged event and a priority list of actions, and providing instructions to a first pump of the hydraulic fracturing fleet equipment to perform the physical action based on the flagged event and the priority list of actions.