The frac: More Art than Science?

I was chatting with an operator about the complexity of it all:  land and mineral rights, the logistics of the spud, reaching  the TD on schedule, the frac itself, the flowback.   And everything that can -  and does - go wrong.   A drilling site with no emergenies is a pipe dream.   Until now.

This image of this frac job below shows how much power converts to pressurizing the hole.   What it doesn't show is the  frac sand (the proppant) and its characteristics.  Shallow well?  White sand might be adequate.   Deep, high pressure well?   Maybe ceramic proppants.  And the mesh size?  20/40, 70/100?  It sure is complicated. 

Nor does this picture show the choke for the initial flowback.   How much water, how much oil, measured by hours, by days.   

What could possibly go wrong?

It's complicated.   Thus, "It's as much art as it is science'.

But more science, more analysis, happening in real time, can paint a less chaotic picture.  The "Internet of Things" provides cross platform connectivity, always measuring and learning, ultimately developing predictability, even in a dynamic environment.   Predictable wear can be scheduled.  A predictable flowback  enhances value to a buyer.   Predictable lets us sleep at night.

While keeping all that expertise -  intellectual property - on-site and secure.  

Therer's a lot of science to grind through.  On premisies AI can do the math, will learn, will compound its learning, and will notice trends before they're emergencies.  Incremental productivity has brought fracking a long way.  Transformational productivity is the next level.