(591e) Handling Microseismic Measurement Uncertainty in Fracture Geometry Control of Hydraulic Fracturing

In hydraulic fracturing, it is important to create fractures with a desired geometry to maximize the oil and gas extraction from unconventional oil and gas reservoirs [1]. In our previous work, we developed a real-time feedback control system for hydraulic fracturing to achieve a desired fracturing geometry by manipulating the proppant pumping schedule; in particular, unmeasurable output and state variables (e.g., average fracture width and propped fracture height) were estimated utilizing available measurements (e.g., fracture length) via a Kalman filter [2, 3]. In hydraulic fracturing, microseismic monitoring (MSM) is the most commonly used technique to access the fracture length, however the associated measurement uncertainty is very high due to the remote nature of hydraulic fracturing taking place at 10,000 feet below ground. Without properly handling this challenge, the high measurement error may lead to incorrect state and output estimation and thereby to poor controller performance. Fortunately, unlike other industrial applications, in hydraulic fracturing the occurrence of measurement depends on the fracturing fluid injection rate at the wellbore. For example, more microseismic events can take place due to increased stress triggered by higher fracturing fluid injection rates; therefore, creating more microseismic events can reduce measurement errors using MSM [4]. This suggests a new state estimation and control framework that utilizes the dependence between the fracturing fluid injection rate (i.e., manipulated input) and measurement error [5].

Motivated by this unique nature of MSM, we propose a novel control framework for measurement uncertainty reduction while simultaneously accomplishing the original control task of achieving the desired fracture geometry in hydraulic fracturing. Initially, a novel high-fidelity model is developed to describe hydraulic fracturing as well as the dependence between the fracturing fluid injection rate and measurement error. Then, open-loop simulation results from the high-fidelity model are used to develop a reduced-order model for a Kalman filter. Then, a model-based feedback control system is proposed using the Kalman filter to achieve both the measurement uncertainty reduction as well as propped fracture geometry control, by manipulating the fracturing fluid pumping schedule which includes the fracturing fluid injection rate and proppant concentration at the wellbore. Closed-loop simulation results are presented to show that the proposed technique outperforms conventional techniques employed in the shale reservoirs.

References:

[1] Economides, M.J., Oligney, R.E., & Valkó, P.P. (2002). Unified fracture design. Orsa Press.

[2] Siddhamshetty, P., Kwon, J.S., Liu, S., & Valkó, P.P. (2017). Feedback control of proppant bank heights during hydraulic fracturing for enhanced productivity in shale formations. AIChE Journal, 64(05), 1638-1650.

[3] Siddhamshetty, P., Wu, K., Kwon, J.S., 2019. Modeling and control of proppant distribution of multi-stage hydraulic fracturing in horizontal shale wells. Ind. & Eng. Chem. Res., 58, 3159-3169.

[4] Maxwell, S.C., Rutledge, J., Jones, R., & Fehler, M. (2010). Petroleum reservoir characterization using downhole microseismic monitoring. Geophysics, 75(5), 75A129-75A137.

[5] Sun, Z., Gu, Q., & Dykstra, J. (2016, July). Uncertainty reduction of hydraulic fracturing process. In 2016 American Control Conference (ACC), 2135-2141.