TY - JOUR
T1 - An empirical model for calculating uniaxial compressive strength of oil well cements from ultrasonic pulse transit time measurements
AU - Anya, Alexander
AU - Emadi, Hossein
AU - Watson, Marshall
N1 - Funding Information:
The authors would like to thank Jorge Quintana, an undergraduate student at Texas Tech University for his help with running the tests.
Publisher Copyright:
© 2019 Elsevier B.V.
PY - 2019/12
Y1 - 2019/12
N2 - Thirty-five cement samples were tested in an Ultrasonic Cement Analyzer (UCA) at varying curing conditions and for varying curing times, and then uniaxially compressed to failure in a hydraulic press. An attempt was made, using the ultrasonic transit time data measured with a UCA and the uniaxial compressive strength data measured using a hydraulic press, to derive an empirical model capable of accurately predicting uniaxial compressive strength from the ultrasonic transit time. Using the Levenberg-Marquardt algorithm, the experimental data was fit to two models. In addition to the transit time dependence, the first model presented incorporates a term that accounts for the curing time while the second model contains terms that account for both the curing time and cement composition. The results obtained illustrate the complexity of deriving a correlation with general applicability to all well cements. The best performing model based on both curve fit error and prediction performance for the samples tested was the model that accounted only for the curing time dependency of the cement strength development rate in addition to the measured transit time, indicating that the transit time measurement history is sufficient as a tool for characterizing well cement strength development.
AB - Thirty-five cement samples were tested in an Ultrasonic Cement Analyzer (UCA) at varying curing conditions and for varying curing times, and then uniaxially compressed to failure in a hydraulic press. An attempt was made, using the ultrasonic transit time data measured with a UCA and the uniaxial compressive strength data measured using a hydraulic press, to derive an empirical model capable of accurately predicting uniaxial compressive strength from the ultrasonic transit time. Using the Levenberg-Marquardt algorithm, the experimental data was fit to two models. In addition to the transit time dependence, the first model presented incorporates a term that accounts for the curing time while the second model contains terms that account for both the curing time and cement composition. The results obtained illustrate the complexity of deriving a correlation with general applicability to all well cements. The best performing model based on both curve fit error and prediction performance for the samples tested was the model that accounted only for the curing time dependency of the cement strength development rate in addition to the measured transit time, indicating that the transit time measurement history is sufficient as a tool for characterizing well cement strength development.
KW - Correlation
KW - Oil well cement
KW - Ultrasonic cement analyzer
KW - Uniaxial compressive strength
UR - http://www.scopus.com/inward/record.url?scp=85071946688&partnerID=8YFLogxK
U2 - 10.1016/j.petrol.2019.106387
DO - 10.1016/j.petrol.2019.106387
M3 - Article
AN - SCOPUS:85071946688
VL - 183
JO - Journal of Petroleum Science and Engineering
JF - Journal of Petroleum Science and Engineering
SN - 0920-4105
M1 - 106387
ER -