Conversely, the big fracking players were much less concerned with, to be completely frank, safety and well integrity compared to the major players, so moved faster, often with troubling but quite profitable results. (Even so, those companies have invested quite a bit in university research to figure out what’s going on, while my client’s company made a major onshore bet, albeit after he retired from leadership.)
Edit: for example chemical industry starts with laboratory experiments, then pilot plant, only then scale-up.
In modelling oil wells, oil industry engineers typically have lots of variables but far fewer data points (well head pressures, flow rates and hopefully some downhole pressures/temperatures) which makes building a good physics based mathematical model tough. The structure and permeability of the rock matrix is difficult to characterise both before the fracking and after. All situations are different and once you start operations which are expensive, you can’t go back and start again if you make mistakes.
There have been considerable improvements in model theory, sensors/instrumentation and computer power in the past decades. Hopefully this can be harnessed to provide cheaper geothermal power for the future as well as managing oil fields more effectively.