The Stand Under the Array: Mounting Structures and Wind Load for Solar Pump Arrays

The short answer: a solar pump array is a sail planted in the most wind-exposed ground on the farm, and the part that fails in a storm is almost never the panel — it is the frame, the fasteners and the foundation under it. Every solar pump lives on open, unshaded ground because that is where the sun is best, and open, unshaded ground is exactly where design wind pressures peak. This guide covers why modules survive storms their structures do not, the wind arithmetic per module, the corner zones that take two to three times the interior load, the design chain from wind speed to foundation, the geometry that buys resistance cheaply, and the inspections that keep the structure honest before and after storm season. Nine models cited here are verified against the live product database.

The array is a sail, and the pump never sees the storm

Start with the asymmetry that defines this whole topic. The pump hangs underground, in the well, below the frost line and below the wind. The controller sits in its enclosure. The one component of a solar pump system that lives outside, unattended, for twenty-five years is the array — mounted on a frame, on the ground, in the open.

And the ground a pump array wants is the ground wind likes too. Shading rules push arrays away from trees and buildings; the 1.3x sizing rule (panels ≥ 1.3 × pump power, the figure on every spec sheet in the catalogue) means the array is big — a 4/6DSC65-85-380/520-7500-A/D (7500 W, 65 m³/h, 85 m) wants a 9,750 W array, roughly eighteen 550 W-class modules and near 40 square metres of glass. All of that glass stands a metre off the ground on the exposed side of a field, tilted at the sun, doing a convincing impression of an aircraft wing working in reverse.

The wind does not push an array down. It tries to lift it up. That single fact rearranges every design instinct: the load that matters is not weight but suction, the worst of it lands at the perimeter rather than the middle, and the structure’s job is not to carry the array but to keep it from flying.

What actually breaks: the module survives, the structure fails

The module is the strongest link in the chain, and it is certified that way. Under IEC 61215, a module type passes static mechanical loading by taking 2400 Pa of uniform pressure on its front and back — with the front pushed to 5400 Pa in the enhanced test used for snow country — and still delivering at least 95% of its rated power afterwards, without glass breakage, cell cracking or frame separation. Dynamic loading adds ±1000 Pa applied 1000 times to shake out fatigue in the laminations and solder joints. A quality 550 W-class module in a 2.3 m² frame therefore carries a factory certificate covering pressures in the 1000-2400 Pa band — comfortably above the design pressure most inland sites will ever see at module height.

Post-storm inspections in the cyclone belts keep reporting the same pattern. In Bangladesh’s delta, after Bay of Bengal storms: the panels are usually fine — it is the racking underneath that fails, with bent purlins, sheared bolts and piles pulled clean out of soft soil. Once the structure goes, the panels go with it: a 28-kilogram module released from a collapsed rail at storm speed meets the ground like anything else dropped from a metre up, and the glass rarely wins.

So the certification question splits in two. The module’s 2400 Pa rating answers “can the panel take the ride?” — usually yes. The system question is “does anything hold it down when the pressure hits 2000 Pa at the corner?” — and that is answered by rail section, clamp count, fastener grade and embedment depth, none of which appear on a module datasheet.

Wind arithmetic: the force per module, worked out

The working formula takes one line. Dynamic wind pressure is:

q = 0.613 × V² (Pa, with V in m/s)

A 40 m/s storm — a strong tropical cyclone’s edge, or a freak temperate-line-squall gust — produces q = 0.613 × 1600 ≈ 980 Pa. Now put that next to the module’s own weight: a 550 W-class module is roughly 2.3 m² of glass weighing 27-28 kg, so its dead weight presses down at about 120 Pa. A 40 m/s wind sucks upward with eight times the module’s own weight before any shape factor is applied.

Shape factors make it worse, because a tilted panel is not a flat plate. Net pressure coefficients for tilted arrays run near 1.0 for an interior module and from 2.6 to 3.2 for a corner one (ASCE 7-style values, uplift negative). Apply them at the same 40 m/s:

Read the corner row twice. In a 40 m/s storm, one corner module experiences an upward pull of over half a tonne — twenty times what it weighs. No clamp “holds” that in any intuitive sense; what holds is the continuous load path from module frame through rail, post and embedment into soil stiff enough to argue back.

Scale it up to a real pump array and the numbers stop being cute. The flagship 8DSC150-37-380/520-11000-A/D (18500 W, 150 m³/h, 37 m) wants a 24,000 W array — forty-four 550 W-class modules, near 100 square metres of glass on a fenced footprint approaching 200 square metres. In the same 40 m/s storm, the air tries to lift that sail with a total force in the rough neighbourhood of eight tonnes. The stand under it is not furniture. It is the only thing between a working pump and a field full of broken glass.

Zones: corners and edges take two to three times the load

Wind pressure does not distribute evenly across an array. Flow separating around the edges creates suction concentrations at the perimeter, and the pattern is brutal and consistent across every wind code: corner modules face roughly 2-3 times the uplift of interior modules, edge modules about 1.5-2 times. The interior of the array partially shields itself; the boundary takes the hit.

Three practical consequences follow.

First, perimeter clamp spacing must be tighter than interior spacing. A design that specifies one uniform clamp pitch across the whole array — the lazy default — quietly under-builds exactly the modules that fail first. Ask any structure supplier what the clamp spacing is in the corner zones and what each clamp is rated to carry there; if the answer is a single number for the whole array, the zoning was not done.

Second, the array’s corners are where post-storm inspection starts. Movement, lifted clamps and hairline glass cracks show at boundary modules before anywhere else. Mark every clamp position with paint at commissioning — after the first storm, any clamp that has shifted off its mark has moved under load, and the module behind it is suspect.

Third, site the array to shrink its own perimeter exposure where you can. A compact rectangle has less boundary per module than a scattered L-shape; a row layout stepped back from a fence line avoids the turbulence a fence adds; and the hedge, earth bund or building that shelters one side (without shading — keep the 9-11 m clearance rules from the shading guide in mind) cuts the worst-zone pressures for free.

From wind speed to foundation: the design chain

Every honest mounting design is the same five-step chain, and you can check any quote against it:

1. Design wind speed (V). From your country’s structural code, not the weather. Cyclone-belt coasts run to the top of the scale — Australia’s AS/NZS 1170.2 assigns Region C 66 m/s and Region D 80 m/s, and a Category 5 typhoon sustains roughly 70 m/s; inland temperate sites typically sit at 25-35 m/s. Add 15-25% if the site is on a hilltop, a valley mouth or a cliff edge.
2. Velocity pressure (q = 0.613 V²). At 33 m/s, q ≈ 670 Pa. At 66 m/s, q ≈ 2670 Pa — a cyclone-belt array starts with four times the base pressure of an inland one, before zones.
3. Zone coefficient. Interior ~1.0, edge ~1.8, corner ~2.8 (use your local code’s table; the ratios are near-universal).
4. Tributary area. The module area each clamp, post and foundation actually carries — a corner post may carry one and a half modules’ area while an interior post carries one.
5. Member and foundation capacity. Clamp rating ≥ zone pressure × tributary area, with margin; fastener pull-out and post embedment designed for the same number, in your actual soil.

The chain lands on the foundation more often than anywhere else. Standard pile embedment of ~1.2 m becomes 2.1-2.4 m in cyclone zones; soft delta or sandy soils need deeper or larger footings still; and ground screws (helical piles) earn their cost where soil resists but digging is hard. Concrete footings run roughly $40-200 per point depending on size and depth; driven piles are the cheapest foundation where soil accepts them ($0.02-0.04 per watt of array); ballast blocks suit rock or sites where penetration is impossible but are the most variable option ($0.04-0.09 per watt).

Two material notes that matter near coasts: fasteners in cyclone or marine zones should be 316-grade stainless, not 304 — chloride and cyclic loading kill 304 within a few seasons — and hot-dip galvanized steel structure wants a heavy coating (120-150 microns, or a zinc-aluminium coating) within salt air range. The cable and corrosion notes for coastal installations apply to the frame, not just the wiring.

For anything above a two-module kit frame, put a local structural engineer’s eyes on the design. The fee is a rounding error against a thrown array.

Cheap geometry: tilt, height and layout that buy resistance

Before spending a cent on steel section, spend decisions. Geometry changes wind load more cheaply than structure does.

Lower the tilt in cyclone country. Arrays in cyclonic regions typically run 10-15° of tilt instead of the 20-25° an inland array would use — a lower profile cuts wind loading by roughly 30-40% while keeping acceptable yield in latitudes below about 25°, where most cyclone-belt pump markets sit (Bangladesh’s coast, the Philippines, Mozambique, Madagascar). The yield trade-off is real but modest; the structural saving is not. Our tilt-angle guide explains the yield side; in cyclone country, wind pushes the optimal tilt down toward flat, and the desert habit of 15-25° so wind and rain shed sand is compatible with it.

Keep the array low and close to the ground. Pressure grows with height above ground; an array on 0.5 m posts sees meaningfully less suction than the same array on 2 m legs. Tall pole-mount arrays — the anti-theft extreme of six-metre poles with razor wire — buy security at a real structural price and belong only where theft, not wind, is the binding risk. For most sites, a low ground frame behind a fence is the honest answer to both problems: the anti-theft guide covers the envelope.

Long continuous rails beat short spliced ones. Rails that span multiple posts without breaks distribute load across the whole frame; every splice is a hinge waiting for a moment. Doubled fasteners with locking features at connection points prevent the vibration-driven loosening that turns a tight structure into a loose one over three storm seasons.

Let the rows help each other. Parallel rows close together partially shield one another’s interiors — this is why utility arrays pack rows tight. Do not overstate it: the leading row and every perimeter take full zone loads regardless, and the shading rules (2-3 panel heights of row spacing at winter solstice, from the shading guide) still govern. But within the spacing the array needs anyway, keeping rows level and aligned beats scattering modules across the field.

Matching the structure to the pump ladder

What the chain means in practice, tier by tier, using the catalogue’s own 1.3x panel rule:

Two tiers deserve a comment. The small tier fails not from engineering but from improvisation: the single pump that “just leans the two panels against a post for now” and is still leaning two seasons later — a two-module A-frame costs little and is the difference between a bracket and a kite. The top tier fails from under-specification: at forty-four modules the array is site infrastructure, and the same string-design discipline that places the modules electrically needs a structural twin that places the posts, embedment and clamping to survive the site’s own worst wind.

Before the storm and after it: the inspection that keeps the structure honest

Structures loosen. Thermal cycling and vibration from ordinary wind work fasteners loose over months, and the failure at the next storm starts at the loosest bolt. A Pilbara pre-season inspection found 15% of foundation bolts under-torqued — on an array that had never seen a cyclone.

Before each storm season (once a year, more often in the first year):

• Torque-check every structural fastener against the design values; re-tighten, and log which ones moved.
• Walk the perimeter looking for posts that have moved, heaved or settled; check footings for undercutting after heavy rain.
• Confirm every clamp sits on its paint mark; any clamp off its mark has shifted under load.
• Remove everything loose around the array — spare clamps, offcuts, drums, sheet material. Loose objects do much of the damage that gets blamed on wind.
• Check cable management: nothing hanging, chafing or flapping; glands sealed. Horizontal storm rain finds every IP gap — this is where the controller enclosure earns its rating.
• Decide, in writing, who shuts the system down at what warning level.

After a major storm, before restart:

• Perimeter walk first, ground level, daylight: displaced modules, lifted clamps, cracked glass (look at low sun angle), rails level, posts upright.
• Photograph everything before touching anything — insurance and warranty conversations both start here.
• Electrical check: each string’s open-circuit voltage and insulation resistance against the commissioning numbers from your acceptance test; string currents compared against each other on the first clear restart.
• A string that lags its neighbours after a storm has a connection or module problem the wind uncovered — find it before it becomes the next storm’s origin point.

What the structure is worth

Run the money. Racking and mounting hardware runs $0.07-0.20 per watt of array installed globally — for a fixed-tilt ground frame, $0.08-0.12 per watt is the working band. On the 7500 W machines above, that is roughly $600-1,200 of steel and concrete carrying a $3,000-5,000 array — the structure is 10-15% of the system cost, and it is the cheapest insurance in the whole project, because the failure it prevents is the only one that destroys everything at once.

The asymmetry runs the other way too. A pump that fails sits in the well and comes up for repair; a controller that fails replaces for a couple hundred dollars; panels that age slowly degrade the yield a few tenths of a percent a year for decades. But a structure that fails in a storm drops forty-four modules onto a field at once, snaps every rail, tears every cable out of the controller, and leaves a working, undamaged pump standing in a well with nothing to feed it. The pump survives; the system dies. The warranty guide sorts out what each component’s warranty covers — no panel warranty covers glass broken by a rail failure, and no pump warranty covers the days a dead array costs.

So buy the structure like it matters, because it is the one part of a solar pump system where the cheapest option and the correct option are different things. Ask for the calculation, not the slogan. Zone the clamping. Torque on a schedule. Walk the perimeter after every big blow. The array is a sail; the stand under it is what decides whether your pump spends the next twenty-five years pumping.

Related reading: array string design against the controller window, panel tilt angle and the fixed-mount case, shading, bypass diodes and row spacing, the commissioning test that sets your baseline, installation guide and mounting checklist, and what lightning and surge protection expect from your earthing.