A direct (open-loop) system pumps the drinking water itself through the collectors. An indirect (closed-loop) system circulates a propylene glycol mixture through the collectors and hands the heat over through a coil in the tank. Indirect is the standard anywhere frost or hard water occurs: a 40% glycol mix protects the loop to about −20 °C, at the cost of a few percent of yield.
What is the difference between direct and indirect solar circulation?
The difference is what flows through the collector. In a direct system — also called open-loop — the potable water that comes out of the tap is pumped up to the roof, heated in the absorber and returned to the store. In an indirect, closed-loop system the collector circuit is a sealed loop of water mixed with propylene glycol; it never touches the drinking water and gives up its heat through a coil heat exchanger inside the tank.
Everything else follows from that one choice. Direct systems are simpler and slightly more efficient because there is no heat exchanger between sun and store. Indirect systems tolerate frost, hard water and stagnation, which is why they dominate every European market north of the olive line.
When is a direct system the right choice?
In frost-free climates with soft water, and mainly for low-cost or large-volume installations. The U.S. Department of Energy is explicit about the split: direct circulation for non-freezing climates, indirect where a heat-transfer fluid and heat exchanger are needed. With no heat exchanger to cross, a direct system delivers collector heat straight into the store, so the collector runs a few degrees cooler and therefore a few percent more efficiently — a real gain that the collector efficiency curve makes visible.
The risks are equally clear. One hard frost can split an absorber; water above roughly 200 mg/L of hardness will scale the absorber channels at the 60–80 °C the collector reaches daily; and the whole circuit is potable, so every component must be drinking-water approved.
When do you need an indirect system?
Any time the collector can go below 0 °C, any time the mains water is hard or aggressive, and any time the system may stagnate. In central Europe that is effectively always. The glycol loop is small — typically 20–60 litres — so it can be dosed, tested and replaced as a consumable, while the drinking water stays in the tank.
Indirect systems also make stagnation survivable. When the store is full and the pump stops, the collector fluid boils and is pushed into the expansion vessel; a properly designed loop simply refills and restarts when it cools. Do that with potable water in an open-loop system and you have scale, corrosion and a hygiene problem.
Direct, indirect or drain-back: the comparison
| Direct (open loop) | Indirect (closed loop) | Drain-back | |
|---|---|---|---|
| Fluid in collector | Potable water | Water + 33–45% propylene glycol | Water (or glycol), drained when idle |
| Frost protection | None | To about −20 °C at 40% mix | Absolute — collector is empty at rest |
| Hard-water tolerance | Poor above ~200 mg/L | Full | Full |
| Heat exchanger loss | None | 3–8% of yield | Small |
| Maintenance | Descaling | Fluid test every 2 years, change every 5–8 | Fluid test, plus slope integrity |
| Typical market | Mediterranean, Middle East, tropics | Central and northern Europe | Central Europe, self-builders |
What about drain-back as a third option?
A drain-back system is the elegant compromise: the collector loop empties into a small reservoir whenever the pump stops, so the absorber is dry — and unfreezable, unboilable — at night, in a power cut, and during stagnation. The price is discipline. Every pipe must fall continuously back to the reservoir at 2–4%, the pump must be strong enough to lift the fluid to the roof on every start, and the collector must be certified as drain-back capable.
What does each option demand from the tank?
Direct systems need a store that is itself potable-rated and, ideally, protected against scale. Indirect systems need enough exchanger surface: plan roughly 0.15–0.2 m² of coil per m² of collector aperture, positioned in the bottom third of the tank so it meets the coldest water and preserves stratification. An undersized coil is the most common reason a correctly sized collector field underperforms — the heat is generated but cannot get into the water fast enough, so the collector runs hot and its efficiency falls.
Solimpeks builds pumped indirect systems for European frost markets and TSM thermosiphon units for warm-climate markets — the TSM heats domestic water through an AISI 316L exchanger rather than passing it through the collector, and the tank specification — not the collector — is what changes most between the two.
Frequently asked questions
Is a direct solar water heater more efficient than an indirect one?
Slightly. Removing the heat exchanger lets the collector run a few degrees cooler for the same store temperature, typically worth 3–8% of annual yield. That advantage disappears the first winter night in a frost climate, which is why indirect systems are standard in central and northern Europe.
Can I convert a direct system to indirect?
Yes, but it usually means changing the tank. The conversion needs a store with a solar coil (or an external plate exchanger plus a second pump), a sealed expansion vessel sized for stagnation, and a glycol charge. In practice most conversions happen when the cylinder is replaced anyway.
Which glycol concentration should an indirect system use?
Solar-grade propylene glycol at 33–45% by volume, giving frost protection to roughly −15 °C to −25 °C. Never use automotive ethylene glycol: it is toxic, and its inhibitor package is not designed for the 150 °C-plus that a stagnating collector reaches.
Do evacuated tubes change the direct-versus-indirect decision?
They reinforce it. Evacuated tubes reach higher stagnation temperatures than flat plates, so the fluid is stressed harder and potable water in the absorber is a worse idea. Almost all tube systems sold in frost climates are indirect.