Deviated wells are where rod pump simulation stops being optional. In a vertical hole, you can get away with rough approximations - gravity pulls the rod string straight down, contact forces are minimal, and most of the physics is one-dimensional. Add 30 or 45 degrees of inclination and none of that holds. The rod string presses against the tubing wall, friction eats into your available load, and every dogleg becomes a potential failure point.
This is also where the differences between simulation tools actually matter. In a vertical well, most software gives you roughly the same answer. In a deviated well, the quality of your simulation depends on how the tool handles survey interpolation, side load calculations, and rod-tubing contact mechanics. Get those wrong and your rod guide spacing is guesswork.
What makes deviated simulation different
The core challenge is rod-tubing contact. In a deviated wellbore, the rod string does not hang freely - it rests against the low side of the tubing at every point where the wellbore curves. Each contact point generates side loads that create friction, accelerate wear on both the rod body and tubing wall, and change the effective load the surface unit has to carry. The magnitude of those side loads depends on the local curvature of the wellbore, the rod weight per unit length, the inclination angle, and the dynamic forces from the pumping cycle.
Survey resolution is the second factor. Directional surveys are typically shot every 50 to 100 feet, but the actual wellbore trajectory between stations is not a straight line. How a simulator interpolates between survey points determines whether it captures the real curvature or smooths it away. A dogleg spanning 30 to 40 feet can fall entirely between two survey stations spaced at 50 feet, which means the simulator never sees it unless it interpolates at a finer resolution.
The third piece is rod guide placement. Side load analysis is not academic - it directly drives where you place rod guides, how many you use per joint, and whether you need to upgrade from standard nylon to metallic guides in high-load zones. If the simulation underestimates side loads because it missed a dogleg or averaged across too coarse an interval, you end up with rod-on-tubing contact in the field that the model said would not happen.
Step length is the most underrated variable
Most rod pump simulators discretize the wellbore into fixed-length segments and solve the wave equation at each node. The length of those segments - the step length - controls how much detail the simulation can resolve. At 50-ft steps, a typical 8,000-ft well has 160 calculation points. That sounds like enough until you look at where the action is.
Doglegs in horizontal and deviated wells commonly span 20 to 40 feet. A build section might gain 3 to 5 degrees per 100 feet over a 500-ft interval, but within that interval the curvature is not uniform. There are localized bends, and those bends are where the peak side loads occur. At 50-ft steps, the simulator averages across each segment, and a sharp dogleg that spans 30 feet gets diluted into the average curvature of the surrounding 50-ft block. The peak side load at that dogleg might be 2 to 3 times what the 50-ft average shows.
For wells with doglegs above 2 degrees per 100 feet, 50-ft steps are inadequate. That is not a soft recommendation - it is a mathematical limitation. You cannot resolve a feature that is smaller than your grid spacing. This is the same reason reservoir engineers use local grid refinement around wellbores: the physics demands it.
How each tool handles it
RODSTAR-D
Theta/RODSTAR offers two separate products for rod pump simulation. RODSTAR-V handles vertical wells and RODSTAR-D handles deviated wells - they are separate licenses at $420/month each. RODSTAR-D includes directional survey import, side load calculations, and rod guide optimization. It has been the default choice in many large operators for years, partly because Theta has been in this space longer than anyone else. The limitation is the fixed 50-ft step length, which means it carries the resolution constraints discussed above.
SROD
Echometer's SROD handles deviated wells within the same license as vertical wells - no separate purchase required. Version 9.2.0 includes a 3D wellbore viewer that lets you visually inspect the rod string trajectory alongside the directional survey, and it calculates drag and side loads along the wellbore. The 3D visualization is genuinely useful for communicating with field teams who need to understand where the high-contact zones are. SROD's user base tends to overlap with operators already using Echometer's dynamometer hardware, which creates a natural workflow from data acquisition to simulation.
XROD AI
Norris/Dover's XROD AI does not support deviated wells. It is a vertical-only tool. If you are evaluating simulation software specifically for directional or horizontal rod pump applications, XROD AI is not in the conversation. That is worth stating plainly because the product name and marketing suggest broad capabilities, but the simulation engine is limited to vertical wellbore geometry.
PetroBench
PetroBench takes a different approach to the step length problem. Instead of fixing the discretization at 50-ft intervals, the step length is configurable. The platform uses cubic spline interpolation on the directional survey, which means the wellbore trajectory between survey stations follows a smooth curve rather than straight-line segments. You can run the simulation at 10-ft, 5-ft, or even 1-ft steps - up to 100 times the resolution of the 50-ft standard.
The practical effect is that doglegs and localized curvature changes actually show up in the side load profile. A 30-ft dogleg that would be averaged away at 50-ft resolution becomes visible at 10-ft steps, and the peak side load at that location resolves to its true magnitude. That changes rod guide placement decisions - you might find that a specific 200-ft interval needs guides on every joint rather than every third joint, something you would miss entirely at coarser resolution.
Because PetroBench runs in the browser, there is no desktop install or license server to manage. Your team shares the same well models, the same simulation configurations, and the same results. When someone adjusts rod guide spacing based on a side load analysis, the rest of the team sees it immediately.
The resolution question is not going away
As operators push rod pump applications into wells with more deviation, longer laterals, and tighter doglegs, the demands on simulation accuracy will only increase. The tools that can resolve fine-scale wellbore geometry without requiring the engineer to work around fixed discretization limits are the ones that will keep pace with the wells being drilled today.
