The Shadow Budget: What One Fence Post Does to a Pumping Array

The short answer: shade is not a percentage, it is a cliff. Ask a buyer what a shadow does to a solar panel and the answer is almost always a percentage — some sun is blocked, so some power is lost, proportionally. On a pumping array the truth is steeper: because cells are wired in series, a shadow covering a few percent of one panel can remove a third of that panel’s output — and in the wrong place, it can stop the pump altogether. Pump sites are also fields: fence posts, tank stands, borehole headworks, trees and the neighbours’ new shed all cast moving shadows across exactly the square metres you paid for. This guide puts a price on that shadow budget, explains the bypass diodes that keep a shaded panel from destroying itself, and shows why the controller’s voltage window — not the wattage — decides whether shade costs you a few litres or a whole pumping day. Every model cited is verified against the live product database.

The physics in one paragraph: series cells and one weak cell

Inside a panel, cells are wired in series — the current must pass through every cell in turn, like old Christmas lights. Whatever cell carries the least current sets the current for the whole string. Worse: when a shaded cell blocks the current the others are trying to push, the healthy cells drive voltage across it in reverse, and the shaded cell turns into a resistor that converts their electricity into heat. That is a hot spot. In the worst cases the local temperature reaches 150-200 degrees Celsius — hot enough to soften solder (typical solder alloys melt near 183 degrees), blister the backsheet and permanently damage the module. This is not exotic: module qualification exists precisely because of this failure mode, with the IEC 61215 hot-spot endurance test written to recreate the worst shading case in a laboratory.

The bypass diode: one third at a time

The defence is built into every panel: bypass diodes. A standard 60-cell panel has three of them in the junction box on the back, each wired in parallel with a group of about 20 cells; 72-cell panels use three diodes over about 24 cells each. Under normal sun the diodes sit idle. When a cell group is shaded, its diode switches on within microseconds and routes the string current around the shaded cells — the panel sacrifices that entire group’s output (about one third of its voltage) but the remaining two thirds keep working, and the shaded cells stop cooking.

Three consequences follow, and they are the whole story of shade on a pump array:

The last row matters because it is the only good news. Soft shade — thin dust, light cloud — dims every cell together and the loss is proportional. Hard shade — a leaf, a bird dropping, the edge of a shadow line from a post — kills a whole cell group at a time. That is why a panel that looks “a bit dirty” can be fine while a single dropping the size of a coin cuts its output by a third. Field research on 36-cell panels without protection found a single shaded cell could drag output down by more than 75 percent; with three diodes the damage is capped at about one third per event — still nothing like proportional.

Why pumps feel shade more than grid-tied systems

A grid-tied inverter that loses a third of its string voltage simply delivers less power; the grid absorbs whatever comes. A solar pump controller cannot be so relaxed, because the pump only runs while the array voltage stays inside the controller’s input window. Every controller in the catalogue defines that window explicitly — the HJSN16-9-24-400 expects 30-48 V DC at its input, the HJSN21-15-48-750 expects 60-90 V, the HJSN27-24-110-1500 expects 110-150 V, and the 6DSC65-85-380/520-7500-A/D works its 520-650 V DC bus. Each bypassed third removes roughly 14 V from a typical modern module at maximum power (a 550 W module runs near 42 V, one third of its cells is about 14 V).

Run the arithmetic and the cliff appears:

The same physics in words: a small 24 V or 48 V system has a window barely two modules wide, so one unlucky shadow can drop it out entirely — the pump does not run slower, it stops until the shadow moves. Large high-voltage strings have enough voltage slack that a fully bypassed panel is a percentage, not an event. This is why shading complaints come overwhelmingly from small systems, and why the fix is layout, not a bigger pump.

The mixed-orientation trap: field data from NREL

Pump farms mix orientations for practical reasons — a row along the fence facing east, a second row facing north to dodge the tank stand’s afternoon shadow. A failure analysis by the US National Renewable Energy Laboratory of bypass diodes pulled from a rooftop array explains why that habit is expensive: the string whose modules sat in two different orientations suffered repeated diode failures from long-term thermal overstress, while a parallel string of the same module type on a single uniform plane had none. Non-uniform illumination keeps parts of a panel diode working near their limits for hours every day; NREL researchers rank bypass-diode failure as the highest-risk failure mode in modules, and field surveys have found failure rates from a few percent of modules over 20 years up to nearly half the diodes in one examined fleet. The practical rule costs nothing: one string, one orientation, one shade profile. Panels that must live with a shadow belong on their own string — where their voltage loss only weakens their own pump — or on a different pump.

The pump-site shade inventory

Shade sources on a pumping site are predictable, which means they are auditable:

Two habits turn this table into a budget. First, walk the site at 8:00, noon and 16:00 — once in the season you pump hardest, once in winter — and mark every shadow that crosses a module. Second, treat a shadow that touches the same module daily as a permanent derate, not a curiosity: a panel that loses one third of its mornings loses roughly one ninth of its year, and on a small system it may be taking the whole string below the window on those mornings.

Ten minutes with a multimeter: the array audit

Because shorted diodes are permanent and silent, the array deserves one audit a year, in full sun:

A panel reading one third low should leave the string: re-string the survivors and put the weak panel on the shelf or the spares line. While you are there, wash off the droppings and check that no leaf debris is wedged along the frame’s lower edge — our desert guide covers the full cleaning schedule for dusty sites, and the same hard-shade logic applies everywhere.

Design rules that cost nothing

1. One orientation per string. The NREL failure pattern is the strongest argument: mixed-facing strings age their diodes faster and fail unpredictably. Split the array by orientation into separate strings feeding the same controller in parallel, or feed a second controller.
2. Keep the daily shadow map off the strings. Shift rows a few metres away from the fence line before buying anything; set row spacing at 2-3 panel heights so winter self-shading never lands on the next row — the arithmetic is in our tilt-angle guide.
3. Orient modules in landscape where hardware allows, so a rising or falling shadow band crosses one cell group at a time instead of sweeping across all three substrings at once. Half-cut cell modules divide each substring again and tolerate narrow shadow bands noticeably better.
4. Group the unavoidably shaded panels together on one string, sized so that even with one panel bypassed the string stays inside the window.
5. Trim on a schedule, not on a complaint. A tree that shades the array for two hours every afternoon will win a slow war against a fixed array; cut it in the season you install, not the season it annoys you.

When the window closes: sizing shade into the system

For a new purchase, shade belongs in the specification, not the troubleshooting:

• Small systems (1-2 modules, 24-48 V pumps such as the 3DPC5-45-48-500 or 3DSC6-60-48-750): the window has no slack. Choose the array site for zero daily shade or accept that morning shadows stop the pump until they pass. Two panels is the whole architecture — every watt and every volt matters.
• Mid systems (4+ modules, 110 V pumps such as the 4DSC15-45-110-750 or 4DSC9.5-90-110-1500): there is room to survive one bypassed panel, but a permanently shaded panel is still a permanent one-third array loss — re-string it out.
• Large systems (520 V class, 6DSC65-85-380/520-7500-A/D): the 130 V of window slack absorbs multiple full-panel bypasses; shading management here is about hot spots and cleanliness, not stopping.
• The A/D hedge: wide-voltage A/D variants accept far broader inputs (for example the HJSU28-30-300-2200-A/D(HV) takes 100-350 V DC on top of 85-280 V AC), so a shade-depressed morning can fall back to AC or a generator instead of a stopped pump. That is a resilience purchase, though — the daily water still comes from the array, and the array still wants full sun.

And the margin rule still stands: size the array at 1.3 times pump power as a floor. The margin absorbs heat, dust haze and ageing — it does not buy back bypassed thirds, which is exactly why the layout rules above matter more than an extra panel.

The pre-order checklist

1. ☐ Site walked at 8:00 / noon / 16:00 in the heavy-pumping season; every shadow crossing a module marked on a sketch.
2. ☐ Tree line measured against array growth — a clear site today, three seasons of growth later, is not.
3. ☐ One orientation per string — no mixed-facing panels on a shared string, ever.
4. ☐ Row spacing at 2-3 panel heights for tilted rows; winter shadow never lands on the next row.
5. ☐ Unavoidable-shade panels grouped on their own string, window arithmetic checked with one panel bypassed.
6. ☐ Annual full-sun open-circuit voltage audit scheduled — three readings per panel, ten minutes for a dozen modules.
7. ☐ Cleaning kit at the site — droppings and leaf debris are hard shade and get removed promptly, not at the next visit.

Related reading: panel tilt angle, array string design against the controller window, cleaning and soiling in desert regions, and the maintenance routine that catches a weak panel early.