Optimization of an in-situ polymerized and crosslinked hydrogel formulation for lost circulation control: Dataset and Analysis

Published: 19 October 2021| Version 1 | DOI: 10.17632/mrrr7rk5pr.1
Karl Jan Clinckspoor,


Lost circulation is a recurring problem during well drilling. This work proposed a new gel formulation, and optimizes its composition using a fractional design of experiments. The gel was analyzed using rheology and nuclear magnetic resonance (NMR). The responses were the maximum elastic modulus (G'max), plateau G', gelation time according to the Hill5 model, gel stability (%), gelation time obtained from NMR and the presence of syneresis, also obtained from NMR. Models were created considering the main effects and first order interactions, and showed a very high coefficient of determination, over 0.95. A more detailed ANOVA can be found in the paper. In general, pH and amount of acrylic acid are the most significant effects, but the amount of ammonium persulfate (APS), chromium (III) salt and carboxymethylcellulose are also relevant. The rheological properties match or are higher than those from the literature, showing this formulation is a good candidate for lost circulation control. A set of optimal gel compositions can be obtained from the data, depending on the desired characteristics of the gel, such as its gelation time and viscoelastic properties. These can be used for further tests under high temperatures and pressures, such as a breakthrough pressure test.


Steps to reproduce

Each experiment (1-19) contains two folders with the rheology and NMR data. The rheology data obtained from the rheometer is in the .xml file and contains three sections, labeled (1|..., 2|..., 3|...). 1| is a time sweep, 2|... is a frequency sweep and 3|... is a stress sweep. The Hill5 equation was fitted to the time sweep, and that is available in the .opj files, which can be opened in OriginLab OriginPro 9.0. In the NMR folder, the CPMG experiment output data is available as .dps files. The monoexponential fits are available in the .mdt file. The T2 distributions are in the .txt files that contain the word CONTIN in the file name. "Cropped" in the file name refers to files that had its first initial points removed, to aid in the inversion. The pdf files show the optimized parameters obtained from the equipment. The file names contain their timestamps, and this was used to track the elapsed time. The "simple gel" folder contains a data set used to build a figure showing syneresis. Equipment data analysis was performed using OriginLab OriginPro 9.0 and Microsoft Excel. These files are available together with the data in the "Data.zip" file. A Jupyter notebook containing the DOE analysis and its conda environment specification can be found in the "DOE Analysis.zip". Python 3.9.2 was used to run the jupyter notebook.


Universidade Estadual de Campinas, Petrobras


Nuclear Magnetic Resonance, Petroleum Engineering, Rheology, Loss of Circulation