One Size Up: The Most Expensive Advice in Pump Sizing

The short answer: add one size for safety is the most expensive advice in pump buying, because on a solar pump every extra watt of pump demands an extra 1.3 watts of panels. On a diesel pump, an oversized engine just sips a little more fuel at part load. On a solar pump, the surplus is invoiced twice — the pump body, and the array that must feed it — and the extra capacity does its worst work at noon, the very hours the whole system lives on. This guide covers why the instinct fails here, what an oversized pump actually does to a shallow well, where the money really goes, the close cousins of oversizing (wrong diameter, A/D machines on tiny arrays), the genuinely correct ways to buy headroom, and the tank that usually does the job better. Nine models cited here are verified against the live product database.

Why one size up costs double on a solar pump

The rule arrives with total confidence, usually from someone who has never paid a panel: buy one size bigger, you can never have too much water. On a mains or diesel pump it is almost free advice — the surplus idles at part load, sipping a little more fuel than strictly necessary. On a solar pump the same advice passes through the catalogue’s own arithmetic and comes back doubled. Every model page prints the rule: array ≥ 1.3 × pump power — the derivation is in the watts guide. Take one rung up the ladder, from a 750 W machine to a 1,500 W one, and the array requirement jumps from 975 W to 1,950 W — double the modules, double the mounting rail, double the cable, double the roof or the fence line they occupy.

And the surplus does not even work when you want it to. A pump array lives and dies by its noon window; every watt of panel capacity bought for a pump that spends its day below its own rating is a watt that had a job waiting at high noon and never got hired.

What an oversized pump actually does at noon

The mental image most buyers carry is a safety margin: the pump loafs along, saving itself for the day you need the full flow. The reality is the opposite, and it comes from the pump curve. Power draw follows the flow: a centrifugal pump moved to the right on its curve — more flow, less head per stage — pulls more power, not less. An oversized pump on an adequate array runs flat-out all day. It does not loaf. The surplus you bought never shows up as gentler operation; it shows up as a pump working harder against your well.

The solar twist makes it sharper. From sunrise to about ten in the morning the array cannot feed the full rating anyway, so the pump runs part-speed — harmless, even pleasant. Then noon arrives and the system goes to work. Whatever the well can supply at that hour is what the pump tries to take. If the pump is bigger than the aquifer, noon is precisely when it digs the drawdown cone deepest, drops the water level toward the intake, and either trips the dry-run protection into its retry cycle or — if no probe was fitted — ingests air and air-locks. The system is strongest at the exact hour the oversized pump does its most damage.

The shallow-well failure mode, step by step

Every well has a yield: the rate at which the aquifer refills the bore. Pump faster than the yield for long enough and the water level falls — first to the top of the pump, then past it. The sequence is predictable:

1. First days: it looks like a success. Flow is spectacular, the tank fills by lunch, neighbours come to look.
2. Within weeks: the afternoon dip. Levels fall through the pumping hours and recover overnight. The dip deepens as the season dries the soil.
3. Then the cycling. Dry-run protection starts tripping on hot afternoons; the retry timer runs the pump in bursts. Each cycle re-fills the pipe, so every restart pays the starting surge again — against a declining level.
4. Finally the sediments. A falling level stirs the bottom silt; a pump that was sold for clean water starts eating sand, and the abrasion begins that ends in seized impellers.

Deep boreholes with real aquifer head usually shrug this off. Shallow wells, river-side shallow bores and low-yield holes do not — and those are disproportionately the wells of smallholders, the buyers to whom one size up sounds most sensible because the price gap looks smallest in percentage terms.

Where the money actually goes

The catalogue’s arithmetic is unsentimental. Every extra watt of pump wants 1.3 watts of panel, and the bill lands in four places:

• Panels. The biggest line. Upgrading a 750 W build to 1,500 W adds a full second array — racking, rails, fasteners, the works.
• Controller headroom. The MPPT must digest the array’s open-circuit voltage and current; the window table has to be checked against the new string both at cold dawn and hot noon.
• Cable and protection. Heavier copper for the longer runs, and the string fuses sized to a new Isc.
• Mounting footprint. Twice the modules need twice the fence or roof — and on a smallholding, that land had another job.

Run the honest comparison at one rung of the ladder. A 3DPC3.8-95-48-750 — 750 W, 3.5 m³/h, 95 m head — wants 975 W of array. One size up, the 4DSC9.5-90-110-1500 — 1,500 W, 9.5 m³/h, 90 m head — wants 1,950 W. The pump price roughly doubles, the array doubles, the controller moves up a voltage class from 48 V to 110 V, and the yard gains a second fence of panels for capacity that will, on most days, spend noon fighting the water table.

The two close cousins of oversizing

The same instinct wears other clothes, and both fail the same way.

The wrong-diameter pump. A 4-inch pump is bought where a 2-inch would do, because the price gap looks small and the 4-inch carries a bigger number. The surplus becomes idle iron: mass hanging on a cable, drag in the bore, and nothing running at its design point. The catalogue’s small-end workhorses — the 2DPC1.5-35-24-200 (200 W, 1.5 m³/h, 35 m) and its 3-inch sibling 3DPC5-45-48-500 (500 W, 5 m³/h, 45 m) — exist precisely so that garden, small-tank and shallow-duty buyers do not have to reach for iron they will never use.

The A/D machine on a DC-only array. An AC/DC hybrid such as the 4DSC16-70-110-1500-A/D carries dual inputs and a hybrid controller, bought with a pure DC array that never uses the AC side. The machine runs; the capability you paid for sleeps. An A/D machine earns its price the day the grid arrives or a generator plugs in — if that day is not on your calendar, the DC-only equivalent is the cheaper tool that does the same noon work.

The headroom that actually pays: the A/D path

There is one version of buy-now-expand-later that genuinely works, and the catalogue builds it in. The A/D machines’ controller accepts both DC and AC input, so the array gets sized for today and the AC side sleeps until needed. When the herd doubles or the grid finally reaches the village, a generator or mains line plugs into the existing pump — no second pump body, no controller swap, no downtime.

The 4DSC16-70-110-1500-A/D is exactly this shape of decision at the small-farm tier: 1,500 W DC on the array, AC input waiting in the same box. At the village tier the same logic scales through 6DSC65-85-380/520-7500-A/D (7,500 W, 65 m³/h, 85 m) and up to the 8DSC150-37-380/520-11000-A/D (18,500 W, 150 m³/h, 37 m) — machines whose AC input is a duty-standby argument in itself. Expansion path in the box, not bolted on after the fact.

When oversizing is genuinely correct

Three honest cases, and only three:

1. A measured, planned expansion. The pivot gets a second circle next season; the herd grows from forty to eighty head by a date on the calendar. Buying the next rung now — say stepping from the 4DSC15-45-110-750 to a machine rated for the expanded draw — can be right, provided the noon surplus in the meantime is banked as water (next section) and the well yield supports the end state, not just the purchase.
2. The A/D path above, where the surplus capacity is AC input that literally sleeps until called.
3. The deep-well floor. Sometimes the next size up is the only model in the catalogue that reaches your head at all — a 270 m draw sends you to the 6DSC36-270-380/520-15000-A/D (15,000 W, 36 m³/h, 270 m) whether or not you need its flow. There the surplus is the price of admission, and the fix is not a smaller pump but planning the array and the storage around it.

What never justifies oversizing: a vague feeling that bigger is safer, a salesman’s one-size-up nudge, or last year’s plan that never happened.

The upgrade that beats a bigger pump: the tank

Here is the quiet arithmetic that closes most of these conversations. Storage costs per litre, roughly, a tenth of what panel capacity costs per litre pumped — and unlike panels, a tank never degrades, never needs cleaning, and works at midnight.

Compare two ways to add water security to a farm whose pump runs dry by mid-afternoon:

• Bigger pump: one rung up the ladder, a doubled array, a new controller class, and a well now being pumped faster than before — with a failure mode (Section 3) that did not exist yesterday.
• Bigger tank: one more day of demand stored as water. The pump keeps its own pace, fills the tank on solar time, and the farm draws from storage through the evening and the cloudy spell. The pump runs longer at its best hour instead of harder at its worst one.

The tank also answers the case the bigger pump cannot: cloudy weather. A larger pump on a dim day produces a slightly larger trickle; a tank filled yesterday produces yesterday’s full noon. The battery-vs-tank guide runs the same comparison against batteries and the tank wins there too.

The check that ends the argument: measure the well’s recovery. Pump it hard for an hour, then watch how fast the level climbs back. If the recovery rate in litres per hour is smaller than the pump’s rated flow, no pump size fixes the physics — storage or demand management is the only honest answer, and the next-size-up quote in your inbox is the most expensive piece of paper on the farm.

The decision checklist

Before any one-size-up purchase, five checks in this order:

1. ☐ Write down the measured daily demand in litres — not the worst day ever, the design day from the sizing guide.
2. ☐ Measure the well’s recovery rate: pump hard one hour, time the refill. Pump flow must stay under yield plus the tank’s buffering share.
3. ☐ Price the full rung: pump + 1.3× panels + controller class + cable + racking. Compare against tank capacity for the same money.
4. ☐ If the expansion is real, schedule the A/D path: 4DSC16-70-110-1500-A/D tier and up, sized on DC today, AC waiting.
5. ☐ Re-run the array’s three tests — watts, cold Voc, hot Vmp — on the bigger string before signing anything.

The pre-quote checklist, one line: the well sets the ceiling, the tank buys the margin, and the pump should be the smallest machine that empties yesterday’s sun into tomorrow’s morning.

The full ladder of machines behind the numbers in this guide is laid out in the model ladder guide, and the village-scale tier — where the A/D expansion path earns its keep — is the subject of the village water supply guide.