Single-Axis Trackers, Done Properly: TMY + Full Loss-Chain Simulation in IST PVSolar Simulator 9.0.10
A single-axis tracker doesn't have a tilt angle — it has a different tilt angle every few minutes, all day, every day of the year. That's precisely what makes trackers hard to simulate honestly a tool that reduces a tracker to one "effective tilt" number is throwing away the exact physics that makes tracking valuable — and the exact physics that makes backtracking necessary. IST PVSolar Simulator 9.0.10's SAT module doesn't take that shortcut. It re-runs the entire hourly physics engine with a genuinely time-varying plane of array, driven by the same reference tracking algorithm used in the solar industry's standard tools.
The tracker angle isn't bolted onto a separate model — it drives the same physics engine
This is the part that matters most. Once the rotation angle is computed for a timestep, it's converted directly into that step's instantaneous surface tilt and surface azimuth:
- Surface tilt = the absolute value of the rotation angle
- Surface azimuth = the tracker's axis azimuth, offset ±90° depending on which way the row is currently facing
Those two numbers are then fed into exactly the same transposition function a fixed-tilt array uses. There is no separate, simplified "tracker mode" calculation running in parallel — the tracker's continuously changing orientation flows straight into the full simulation pipeline Perez beam/sky-diffuse/ground-reflected transposition, angle-of-incidence loss, the single/two-diode I-V model, the IEC 61724-1 thermal model, spectral correction, the inverter's real efficiency curve, and — since these were added to the same engine — the per-timestep MPPT voltage-window check and the far-horizon beam-cutoff/diffuse-loss model all apply automatically to the tracker's instantaneous plane of array, with zero duplicate code paths to get out of sync.
A dedicated loss chain for what's specific to trackers
On top of that shared physics core, the SAT module adds a loss chain for the effects that are genuinely tracker-specific and don't apply to a fixed array:
- Soiling and snow coverage, tracked separately from the main array's soiling assumptions, since tracker rows often accumulate dust differently than a fixed tilt
- Mechanical availability (default 99%) — accounting for motor/controller downtime, distinct from inverter or grid availability
- DC wiring loss specific to the tracker's longer, moving cable runs
- Tracker parasitic power (default 8 W per tracker) — the motor and controller's own electricity draw, subtracted from net output
- Module degradation, tracked the same way as fixed arrays but reported alongside the tracker-specific figures
Each of these shows up in its own waterfall breakdown and monthly table, so the tracker's gain over a fixed array is reported net of the costs that come with the moving parts — not just the raw cosine-gain headline number.
A fast preview, then a real re-simulation — not a shortcut masquerading as one
Configuring the tracker (axis azimuth, max rotation, GCR, module width, hub height) and clicking Calculate gives an instant client-side preview: estimated annual H_POA, a monthly gain-vs-fixed table, a POA comparison chart, and a tilt-angle-over-time chart. That's deliberately framed as a preview, not the final answer.
Clicking Apply to Simulation doesn't just bake that preview number into the results — it queues a genuine per-timestep tracker re-simulation to run on the next full simulation, and if the project is set to Monthly resolution, it automatically switches to Hourly, because a tracker's rotation angle is only meaningful at sub-daily resolution. The applied "SAT Tracker Gain (vs Fixed)" loss-table row is explicitly labeled as the fast preview estimate, with the real engine-computed figure arriving once the full hourly run completes — an honest distinction between a quick estimate and the actual physics-based result, rather than presenting one as the other.
Why this design matters
A tracker's entire value proposition — and its entire risk of self-shading loss — lives in the minute-by-minute trade-off backtracking makes between chasing the sun and avoiding row-to-row shadow. Collapsing that into a single "average effective tilt" erases exactly the calculation that determines whether a given GCR and row spacing actually deliver the tracking gain a developer is counting on. By running the real NREL/pvlib-standard rotation and backtracking equations through the same hourly physics engine used for every other loss in the plant — TMY weather, real diode I-V curves, real thermal response, real inverter efficiency, and now the horizon and MPPT-window checks too — IST PVSolar Simulator's SAT module gives a tracker the same rigor a fixed-tilt array gets, instead of treating it as an afterthought bolt-on.