"100x wellbore resolution" sounds like a marketing number. It's not. It's a specific change in how many points along the wellbore the solver uses to calculate forces on your rod string - and it changes what you see in side load output, stress profiles, and guide placement recommendations. Here's what it actually means in terms of simulation output.
The math
Take a 6,000-ft deviated well. The step length is the distance between calculation points - locations where the solver explicitly computes axial load, lateral force, stress, and rod-tubing contact.
- 50-ft steps: 120 calculation points along the string
- 10-ft steps: 600 calculation points
- 5-ft steps: 1,200 calculation points
That's the "100x" number. Going from 50-ft to 5-ft steps doesn't change the physics engine. It changes how many locations the engine evaluates. More points means the solver sees the actual wellbore geometry instead of a smoothed-out approximation of it.
What changes in side load output
At coarse resolution, side loads look like broad humps. You can see that lateral force increases somewhere in a general zone, but the shape is soft and the peaks are averaged down. At 50-ft steps, a side load peak that should read 280 lbs at a specific dogleg might show up as 180 lbs spread across a wider interval. It stays below your 200-lb guide threshold. The well looks clean.
At fine resolution, those humps resolve into sharp peaks at specific measured depths. The same well now shows two or three distinct zones where lateral force crosses the guide threshold. Those peaks are where your rod contacts tubing and where your rod fails. The humps just told you "somewhere around here."
We see this constantly in support conversations. An engineer runs a well at the default step length, sees no guide recommendations, and moves on. Same well at 10-ft steps shows three contact points that need protection. The physics didn't change - the visibility did.
What changes in stress profiles
Stress concentrations live at doglegs. A 2-degree/100-ft dogleg in a 7,500-ft well creates a localized stress riser that might span 20-30 ft of measured depth. At 50-ft steps, the solver evaluates one point on either side of that dogleg and interpolates through it. The stress concentration gets smeared out, and the peak value drops.
At 10-ft steps, the solver places multiple evaluation points through the dogleg itself. The actual peak stress resolves. A section that looked safely below the modified Goodman limit at coarse resolution might be above it at fine resolution - not because the stress changed, but because you're finally measuring it where it actually occurs.
This matters for taper design. If you're deciding between Grade D and Grade K at a taper transition, and the Goodman diagram shows 15% margin at coarse resolution, you might not have that margin when you resolve the actual stress peak. Running the design at 10-ft steps before finalizing rod grades is cheap insurance.
What changes in guide placement
Typical rod guide spacing runs 30-60 ft. If your simulation step length is 50 ft, you're placing guides based on one calculation point per guide interval - maybe two if you're lucky. The side load values between those points are interpolated, not calculated. You're designing guide placement on estimated data.
At 10-ft steps, each guide interval contains 3-6 explicitly calculated points. You can see exactly where within the interval the lateral force peaks, and whether the guide should sit at the top, middle, or bottom of that zone. A 30-ft placement error on a rod guide in a high-dogleg section is the difference between protecting the contact point and missing it entirely.
This is where resolution has the most direct operational impact. The simulation tells you "place a guide at 4,230 ft MD" instead of "place a guide somewhere between 4,200 and 4,250 ft." Your rod service company can actually use that.
When it doesn't help
If your directional survey stations are 200 ft apart, running at 10-ft steps gives you 20 calculation points per survey interval - but the wellbore trajectory between those stations is interpolated from the survey data. The solver is computing forces on an interpolated path, not a measured one. You get smoother output, but you're not adding real geometric information.
Resolution can't exceed your input data quality. Dense surveys (30-90 ft station spacing) paired with fine step lengths give you the most accurate picture. Sparse surveys with fine steps give you a precisely computed result on an uncertain path. Know the difference.
Similarly, if you're running a near-vertical well with minimal dogleg severity, 50-ft steps and 10-ft steps will produce nearly identical output. The resolution payoff scales with wellbore complexity. Straight wells don't need it. Deviated wells with multiple doglegs absolutely do.
The compute trade-off
Running at 10-ft steps takes roughly 15 seconds versus about 5 seconds at 50-ft steps for a typical well. That's a cloud compute problem, not an engineering problem. PetroBench runs these on scalable infrastructure, so the extra 10 seconds doesn't block your workflow or consume a local license. You run it, you wait a few seconds, you get a result with 5x the spatial detail.
For sensitivity runs across multiple designs, the time adds up slightly - but you're still talking minutes, not hours. Desktop tools that take 30+ seconds per run at coarse resolution make fine-resolution sweeps impractical. That's a tooling limitation, not a physics one.
Higher resolution doesn't change the physics. It changes whether you see what the physics is actually doing to your rod string.
