Backtracking Algorithm for Solar Trackers — How It Works and Why It Matters

A utility-scale solar project with single-axis trackers can lose 3–8% of annual energy yield if the backtracking algorithm is incorrectly specified or not enabled. On a 100 MW project in Rajasthan generating 220 MU/year, that is 6.6–17.6 MU of lost generation annually — at ₹2.50/kWh tariff, a revenue loss of ₹1.65–4.4 Cr per year for the full project life. The backtracking algorithm is not a software setting to leave at default. It is a design decision that belongs in the tracker procurement document.

Direct answer. The backtracking algorithm for solar trackers is a geometric shade-avoidance control that rotates tracker rows to a sub-optimal tilt angle during low-sun-angle periods (morning and evening) to prevent row-to-row shading. Without backtracking, string-level mismatch losses from shaded cells can reduce output by 30–70% during these periods because bypass diodes activate and cut entire strings. The Backtracking Decision Window framework presented here maps the three controlling variables — morning shadow hour, clipped energy versus shade loss, and row spacing (GCR) — to the optimal backtracking start and stop angle for a given site, enabling engineers to specify and verify tracker performance with precision.

This article serves utility-scale solar developers and EPCs in India preparing tracker procurement documents for SECI auctions, DISCOM tenders, and captive IPP projects. It links to Heaven Designs’ analysis of solar engineering for Indian EPCs complete workflow and the per-MW solar engineering cost breakdown for tracker-related engineering line items.

What Row-to-Row Shading Does Without Backtracking

At low sun angles — early morning and late evening — the shadow cast by one tracker row onto the next is unavoidable on any ground-mount project with ground coverage ratio (GCR) above approximately 0.20. The length of the shadow depends on the sun elevation angle, the module height, and the inter-row spacing.

The energy impact of row-to-row shading is not linear. When a shadow falls across the bottom cells of a module in a portrait-orientation tracker row, it activates the bypass diodes of the affected cell strings. A single shaded cell string cuts the power output of approximately one-third of the module. When the shadow covers cells across multiple modules in the same string in the DC combiner circuit, the mismatch loss compounds — the shaded string pulls down the entire string voltage, and the combiner box current loss can exceed the shadow fraction.

According to NREL’s Bifacial PV System Performance publication 2020, row-to-row shading without backtracking can cause energy losses of 3–8% annually on systems with GCR above 0.35, compared to a perfectly spaced system. The losses are concentrated in the first 1–2 hours after sunrise and the 1–2 hours before sunset, when the sun elevation angle is low and shadows are longest.

The fundamental physics is simple: a row of 4-landscape or 2-portrait modules on a tracker row at typical Indian heights (700–900 mm hub height) casts a shadow that at 10° sun elevation extends to approximately 3.5–5.5 times the module height. At a GCR of 0.40 with 7m inter-row spacing, that shadow reaches the next row every morning until the sun climbs to approximately 15–20° elevation. Backtracking is the solution.

How the Backtracking Algorithm Works

The backtracking algorithm calculates, for each moment of the day, whether the optimal tracker action is to track the true sun position or to rotate to a flatter angle to avoid casting a shadow on the adjacent row. The calculation is purely geometric — it uses the sun position (elevation and azimuth angles) and the known system geometry (module dimensions, tracker row height, inter-row pitch) to compute the maximum tilt angle at which the row does not shade its neighbor.

The algorithm inputs are:

1. Sun position — calculated from GPS coordinates and time of day using an astronomical algorithm (NREL SPA or similar).
2. Row height — the height of the module top edge above ground in the vertical position.
3. Inter-row pitch — the center-to-center distance between tracker rows.
4. Module dimensions — the portrait or landscape module height determines the shadow length at each tilt angle.

The algorithm output is the backtracking tilt angle — the tilt angle that places the module’s shadow edge exactly at the base of the adjacent row. At this tilt angle, no shade falls on the adjacent row, and the tracker is as steep as possible given the constraint.

The algorithm switches between two operating modes:

True sun-tracking mode: The tracker follows the sun’s actual position to maximize the angle of incidence on the module surface. This is the mode used from mid-morning to mid-afternoon when the sun is high enough that row-to-row shading is zero. In this mode, the tracker tilts to 50–60° from horizontal in summer in high-latitude sites, and 30–40° in Indian latitudes.

Backtracking mode: The tracker rotates to the shade-free angle, which is less steep than the true sun position. In backtracking mode, the tracker sacrifices some direct beam irradiance angle (losing some direct normal irradiance) to avoid the much larger mismatch loss from string-level shading. The net yield effect is strongly positive — even though the tracker is not pointed directly at the sun, the elimination of mismatch losses far outweighs the cosine loss from the sub-optimal angle.

The Backtracking Decision Window Framework

The Backtracking Decision Window maps the three controlling variables — morning shadow hour, clipped energy versus shade loss, and GCR — to the optimal backtracking start and stop angle for a given site.

Apply this framework in the PVsyst simulation setup: enter the correct GCR, enable the backtracking shading model (PVsyst 7.4 calls this “shade limitation by backtracking angle”), and verify that the simulated yield with backtracking exceeds the simulated yield without it by 3–6% on a GCR 0.35–0.45 system. If the gap is less than 2%, the backtracking settings may not be correctly entered.

Impact on Annual Energy Yield: PVsyst Modeling Results

The energy yield benefit of backtracking has been extensively studied. According to NREL’s Single-Axis Tracking Performance study, a system with GCR of 0.40 gains 3.5–5.5% additional annual energy yield with backtracking enabled compared to the same system without it.

GCR	Annual yield without backtracking (relative)	Annual yield with backtracking (relative)	Backtracking gain
0.25	100%	101.0%	+1.0%
0.30	100%	102.5%	+2.5%
0.35	100%	103.8%	+3.8%
0.40	100%	105.0%	+5.0%
0.45	100%	106.2%	+6.2%
0.50	100%	107.5%	+7.5%

Source: NREL Single-Axis Tracking Simulation Analysis, adapted for Indian latitudes (22–26°N).

In PVsyst 7.4, backtracking modeling is enabled in the “Orientation” tab of the system definition, under “Tracking system parameters.” The key settings are:

• Backtracking enabled: yes
• Row pitch: must match the as-designed inter-row spacing (center-to-center)
• Limit angle: the maximum tracker rotation angle (typically ±60° or ±55° for most tracker models)
• Ground coverage ratio: auto-calculated from row pitch and module dimensions — verify this matches your actual layout

A common PVsyst modeling error is to enable backtracking in the simulation but set an incorrect row pitch, resulting in a simulated GCR that does not match the design GCR. The yield error from a 0.5m row pitch error on a 0.40 GCR system is approximately 0.5–1% annually — small in percentage terms but meaningful in revenue over a 25-year project life.

GCR and Backtracking Interaction

GCR and backtracking are the two most tightly coupled design variables in single-axis tracker system design. Changing GCR without re-evaluating the backtracking strategy introduces yield modeling errors that can compromise the bankability of the PVsyst report.

The interaction works in both directions:

Higher GCR with backtracking: Increasing GCR from 0.35 to 0.45 increases land utilization by 29%, but also increases the duration of the morning/evening backtracking window. The annual yield per MW decreases slightly (because the tracker spends more time in sub-optimal backtracking mode), but the annual yield per hectare increases substantially. For Indian utility-scale projects where land cost is significant, higher GCR with backtracking is usually the optimal design.

Lower GCR without backtracking: Spacing tracker rows widely enough to eliminate inter-row shading entirely removes the need for backtracking, but wastes land. At GCR = 0.25 in Indian latitudes, the system achieves zero inter-row shading — but uses 80% more land than a GCR = 0.45 system for the same installed capacity. This trade-off is never justified in Indian utility-scale projects.

The interaction between GCR, backtracking, and bifacial gain adds a third dimension to this analysis. Bifacial modules mounted on trackers gain rear-side irradiance from ground-reflected light. Backtracking reduces the module tilt in morning and evening, which changes the sky-view factor for the rear side. NREL’s bifacial tracking simulation work shows that bifacial gain on a backtracking system is slightly lower than on a true-tracking system (because the reduced tilt during backtracking increases the rear-side ground shading). This effect is small (0.1–0.3% annual bifacial gain reduction) but should be modeled explicitly in PVsyst 7.4 for bankable reports.

How Tracker Vendors Implement Backtracking

All major tracker vendors offer backtracking as a standard feature, but the implementation details differ in ways that matter for energy yield.

Nextracker (NX Horizon): Nextracker’s TrueCapture system uses an AI-based algorithm that adapts backtracking to measured irradiance conditions, not just astronomical calculations. TrueCapture claims 2–6% additional yield versus standard astronomical backtracking by optimizing the transition between backtracking and true-tracking based on diffuse-versus-direct irradiance ratio. For sites with high diffuse irradiance (coastal zones, monsoon-affected sites), TrueCapture’s adaptive approach is particularly valuable. For standard arid sites in Rajasthan, the incremental gain over standard backtracking is 0.5–1.5%.

Array Technologies (DuraTrack): Array Tech implements standard astronomical backtracking with an option for their SmartTrack adaptive algorithm, which adjusts backtracking based on irradiance sensor feedback. SmartTrack is particularly effective on sites with terrain variation or partial cloud cover. Array Tech is one of the most widely deployed tracker brands in India; their backtracking configuration requires the installer to enter the correct pitch and limit angle in the field controller during commissioning — a step that is sometimes skipped by commissioning teams.

Soltec (SF7): Soltec implements a terrain-adaptive backtracking algorithm called “Self-Learning Backtracking” that adjusts for longitudinal (north-south) slope. For Indian sites in Karnataka and Tamil Nadu with rolling terrain, this feature prevents the yield loss that standard backtracking algorithms cause on sloped sites.

PVHardware, Arctech, and other Chinese tracker OEMs: These brands, widely used in Indian utility-scale projects for cost reasons, implement standard astronomical backtracking. The implementation quality varies — some controllers have known firmware issues where the backtracking angle calculation drifts over time if GPS synchronization is lost. Specify firmware version requirements and GPS synchronization backup in the procurement document.

What to Specify in a Tracker Procurement Document

The tracker procurement document (technical specification) for an Indian utility-scale project should include explicit backtracking requirements. Most procurement documents fail to specify these parameters, leaving the tracker vendor to set defaults that may not match the project design.

The minimum backtracking specification items are:

1. GCR or row pitch: state the as-designed inter-row pitch (in meters, center-to-center) and the resulting GCR, calculated from the module dimensions. Require that the tracker controller be configured and commissioned with this exact GCR.

2. Backtracking algorithm type: specify “astronomical backtracking” as the minimum requirement. If adaptive backtracking is required (for diffuse-dominated sites or sloped terrain), name the specific product feature (e.g., “TrueCapture or equivalent adaptive backtracking with irradiance sensor input”).

3. Backtracking verification test: require a Factory Acceptance Test (FAT) or Site Acceptance Test (SAT) that demonstrates correct backtracking operation. The test should record tracker angle versus sun position for at least one full morning and compare the backtracking angle curve against the theoretical backtracking angle calculated from the specified GCR.

4. GPS synchronization and backup: require that the tracker controller maintains accurate sun position calculation in the event of GPS signal loss for at least 72 hours using an internal clock and last known position. Require a manual re-sync procedure be documented in the O&M manual.

5. Terrain-adaptive backtracking (if applicable): for sites with north-south slope exceeding 1%, require explicit terrain-adaptive backtracking capability and a demonstration that the algorithm accounts for the as-built terrain profile.

6. Commissioning verification: require that the commissioning report include a comparison of the as-commissioned GCR parameter versus the design GCR, signed by the tracker manufacturer’s commissioning engineer.

How Heaven Designs Helps with Tracker Design and Yield Modeling

Single-axis tracker system design involves multiple interdependent decisions: GCR selection, row pitch optimization, backtracking specification, terrain grading design, and PVsyst modeling with correct backtracking parameters. Errors in any of these propagate into the yield report and, for bankable projects, into the IREDA or PFC technical review.

• Solar Ground Mount Design — complete tracker system design including GCR optimization, row spacing layout, terrain grading plan, and bankable PVsyst simulation with backtracking enabled per IEA-PVPS Task 13 methodology.
• Solar Civil and Structural Engineering — pile design and structural calculations for tracker foundations, pile pull-out test specifications, and grading design for tracker-compatible terrain profiles.
• MW-Scale PMC — owner’s engineer services for tracker commissioning verification, including backtracking parameter audit and SAT documentation review.
• Download a sample tracker design package — including layout, PVsyst settings, and tracker specification extracts from a real project.

For a tracker system design scope and a yield analysis for your specific GCR and site, contact us.

FAQ

What is the difference between backtracking and true sun tracking?

True sun tracking means the tracker rotates continuously to face the module surface directly at the sun, maximizing the direct normal irradiance (DNI) incident on the module. Backtracking is a modification of this — during periods when true tracking would cause one row to shade the next, the tracker instead rotates to a flatter angle that eliminates inter-row shading at the cost of a slightly non-optimal sun angle. True tracking maximizes instantaneous irradiance at each moment; backtracking maximizes total daily energy yield by eliminating the far larger mismatch loss that shading would cause during the low-sun periods.

Does backtracking work for bifacial modules on trackers?

Yes. Backtracking is fully compatible with bifacial modules and is modeled correctly in PVsyst 7.4’s bifacial simulation module. The rear-side gain of bifacial modules during backtracking mode is slightly lower than during true-tracking mode, because the reduced tilt angle changes the sky-view factor of the rear surface. According to NREL’s bifacial tracking analysis, the bifacial gain reduction during backtracking is typically 0.1–0.3% annually — small enough that it does not change the design conclusion that backtracking is beneficial at GCR above 0.30.

Do systems with MLPE (micro-inverters or DC optimizers) need backtracking?

Systems with module-level power electronics (MLPE) — micro-inverters or DC optimizers at each module — partially mitigate row-to-row shading mismatch losses because each module operates independently. On an MLPE-equipped tracker system, the bypass diode mismatch loss from inter-row shading is eliminated at the module level. This means the yield loss from operating in true-tracking mode without backtracking is lower than on a conventional central-inverter or string-inverter system. However, backtracking still provides a yield benefit on MLPE systems by eliminating the cosine loss from low-angle DNI on shaded cells. The net backtracking benefit on MLPE systems is approximately 0.5–1.5% annually, versus 3–8% on conventional string-inverter systems.

How does soiling interact with backtracking?

Soiling and backtracking interact in a subtle but important way. When a tracker row is in backtracking mode (tilted toward horizontal), dust and soiling accumulate more uniformly across the module surface than when the module is at a steep tracking angle. This means soiling losses on backtracking tracker systems are slightly higher than on fixed-tilt systems during the morning and evening backtracking periods. For high-soiling sites in Rajasthan and Gujarat, model an additional 0.1–0.3% annual soiling loss on tracker systems compared to the standard soiling assumption. This effect is small but should be documented in the PVsyst report for bankability.

What causes backtracking to fail in the field, and how is it detected?

The most common cause of backtracking failure in the field is GPS signal loss, which causes the tracker controller to fall back to a stowed or default position rather than calculating the correct backtracking angle. Other causes include firmware bugs in the tracker controller that cause the backtracking angle calculation to drift, incorrect GCR configuration entered at commissioning, and physical tracker malfunction (motor failure, drive unit damage) that prevents the row from reaching the backtracking angle. Backtracking failure is detected through SCADA monitoring — a tracker operating in true-tracking mode when backtracking should be active will show anomalously high inverter output in the first and last hours of the day compared to neighboring trackers. Performance ratio monitoring at 15-minute intervals is the recommended detection method.

How is backtracking gain verified in the performance test at COD?

The performance test at Commercial Operation Date (COD) for a tracker project should include a specific backtracking verification: record the tracker angle versus time for the first 2 hours after sunrise on a clear-sky day, and compare against the theoretical backtracking angle curve calculated from the as-built GCR and the recorded sun position. A correctly functioning backtracking system will match the theoretical curve within ±1°. Additionally, compare the measured energy yield during the backtracking window (first 2 hours after sunrise) against the PVsyst simulated yield for the same period — a discrepancy exceeding 5% on a clear-sky day indicates a backtracking configuration issue. See also our article on bid-stage vs IFC-stage engineering for where performance testing requirements belong in the project engineering lifecycle.

What is the backtracking specification for SECI tender projects?

SECI tenders for solar projects with trackers increasingly include tracker performance specifications in the technical schedules. As of 2025, SECI’s standard solar PV project tender documents require that tracker systems include backtracking capability as a minimum, with the GCR parameter configured to match the as-designed row spacing. Some SECI tenders additionally require adaptive backtracking for sites in identified high-diffuse-irradiance zones (coastal and northeastern states). The tracker vendor’s technical datasheet submitted in the bid must confirm backtracking capability, and the commissioning report must confirm GCR configuration. IREDA’s technical review of SECI-backed project financing verifies backtracking specification compliance as part of the tracker section review.