Collector efficiency reference

A solar thermal collector's datasheet headline number — the peak optical efficiency η₀, typically quoted at 0.70–0.82 for a good flat-plate panel — is the maximum possible fraction of incident sunlight converted to useful heat at one favourable instant: sun normal to the glazing, fluid entering at ambient temperature.

The annual delivered fraction — useful heat at the tank, divided by sunlight incident on the panel over a whole year — is much lower because of a cascade of physical losses, each modest on its own, that compound through the year. This page lays the cascade out explicitly, shows the Hottel–Whillier–Bliss equation that drives the biggest term, and lists published references that bracket the real-world range.

If you're trying to size a collector array, this page is the "what fraction of the sun do I actually get?" input to the /resources/sizing calculator. If you're trying to orient an existing array, head to the /resources/tilt-optimizer.

From peak η₀ to annual delivered: the cascade

Typical residential drainback flat-plate, mid-latitude site
Stage	Factor	Running
Peak η₀ (SRCC / Solar Keymark nameplate, ΔT = 0, normal incidence)	0.78	78 %
× Temperature derate (collector hotter than ambient — Hottel–Whillier–Bliss)	0.85	66 %
× Cosine-of-incidence averaged across the day (sun isn't normal except at one moment)	0.85	56 %
× Pump / control losses + morning warm-up of cold pipe	0.93	53 %
× Stagnation periods (tank full, collector idle, no useful collection)	0.95	50 %
× Pipe loss to surroundings (worse for long outdoor runs)	0.93	47 %
× Soiling, dust, partial shading over the year	0.97	45 %

A reasonable starting point for seasonal-average delivered efficiency in the /resources/tilt-optimizer or /resources/sizing calculators:

0.40–0.45 — typical residential drainback flat-plate
0.50–0.55 — well-designed short-pipe / low-delivery-temperature install
0.45–0.55 — evacuated tube in high-ΔT applications (the vacuum jacket gives a much lower temperature derate, but optical η₀ is lower to start with)

Hottel–Whillier–Bliss: the temperature-derate term

The single biggest loss after the headline number is the temperature derate. A collector running hotter than ambient leaks heat out of the same absorber that's collecting it; the hotter it runs, the worse the leak. The standard SRCC / ASHRAE-93 single-line model:

η = η₀ − a₁ · (ΔT / G) − a₂ · (ΔT)² / G

η — instantaneous collection efficiency [0–1]
η₀ — peak optical efficiency (datasheet Y-intercept, dimensionless)
a₁ — first-order heat-loss coefficient, W/(m²·K)
a₂ — second-order coefficient, W/(m²·K²) — small for flat-plate, important for high-ΔT evacuated tube
ΔT — T_collector_inlet − T_ambient, K
G — solar insolation on the collector plane, W/m²

In practice for flat-plate panels at moderate temperatures the second-order term is small and can be dropped, giving the linear form:

η ≈ η₀ − a₁ · (ΔT / G)

Worked example — flat-plate at solar noon

Using a flat-plate panel with η₀ = 0.78 and a₁ ≈ 4 W/(m²·K), operating at ΔT = 30 K (collector 30 °C above ambient) and G = 600 W/m² (good but not peak sun):

η = 0.78 − 4 × 30 / 600
  = 0.78 − 0.20
  = 0.58  (instantaneous, solar noon)

So 20 percentage points lost to the temperature-derate term alone, before any of the other cascade losses kick in. The cosine factor, control losses, pipe losses, stagnation, and soiling then drag the annual average down from there.

Crossover: when evacuated tube starts to win

Two collectors with different (η₀, a₁) pairs have their efficiency lines cross at a particular ΔT:

ΔT_crossover = (η₀,FP − η₀,ET) / (a₁,FP − a₁,ET)

Using representative SRCC-certified collector parameters — flat-plate AET AE-32 (η₀ ≈ 0.71, a₁ ≈ 4.9 W/(m²·K)) versus a typical evacuated tube (η₀ ≈ 0.46, a₁ ≈ 1.6 W/(m²·K)) — at G ≈ 250 W/m² (80 % full sun), the crossover sits at ΔT ≈ 34 K. For domestic hot-water duty (T_inlet typically 5–40 °C, ΔT modest), flat-plate always wins. Evacuated tube only pays back its higher cost when the operating ΔT is large (winter space heating in a cold climate, process heat above 60 °C, or stagnation-prone undersized stores). Both parameter sets above are published in the public SRCC OG-100 collector database.

Aperture area vs gross frame area: a hidden ET handicap

The η₀ on every collector datasheet is measured against the aperture area — the active absorber surface where sunlight actually lands. For sizing a real install, the metric that matters is delivered energy per square metre of roof area, and the two diverge significantly between collector types:

Flat-plate — the absorber plate fills the frame. Aperture area is typically 90–95 % of the gross frame footprint. Almost the entire roof patch is doing useful work.
Evacuated tube — aperture area is the projected area of the absorber strips inside the vacuum tubes, which sit with visible gaps between them. Aperture area is typically only 55–65 % of the gross frame footprint. The gaps are dead space.

So an ET array with datasheet η₀ ≈ 0.78 (against aperture) delivers closer to η ≈ 0.45–0.50 per m² of roof. A flat-plate at the same aperture η₀ delivers η ≈ 0.70–0.75 per m² of roof. For a given usable roof area, flat-plate substantially wins on absolute delivered kWh — even before the cascade losses below.

SRCC OG-100 and Solar Keymark certification reports list both aperture and gross areas precisely because the headline η is only valid for the smaller aperture figure; the gap is one of the most common sizing pitfalls when comparing collector products.

Other practical disadvantages of evacuated tube

Fragility. Glass tubes break under hail, ice fall, rooftop foot traffic, or fast thermal shock during cold-mains drainback fills. Replacement is per-tube and requires vacuum-integrity inspection.
Snow shedding — and the preheat trick. On a flat-plate collector, once the sun catches even an edge of the panel after a snowfall, the absorber surface beneath the snow starts to warm; a thin water film forms at the glazing-snow interface and the whole snow load slides off under gravity, often clearing in minutes once the process starts. Evacuated tubes can't do this — the vacuum jacket isolates the absorber from the front of the tubes, so the snow-glazing interface never sees the absorber's heat and no melt-film forms. An ET array stays snow-covered until the load shifts mechanically, often weeks of winter performance lost entirely. (Worth noting: overnight, both collector types radiate net heat to the cold sky and actually run below ambient — that's why frost forms on glazing on a still-mild evening. So the snow-shedding mechanism is sun-driven each morning, not residual heat carried from the previous day.)

A drainback flat-plate's controller can also clear heavier or fresher snow loads on demand by running a short preheat cycle — circulating warm tank water through the panels for ~10 minutes creates the same melt-film from behind the snow, no sun-on-edge required. Drainback makes this nearly free because the tank water is the circulating fluid; closed glycol systems can do the same trick via the loop heat-exchanger, just with one extra step to pre-warm the glycol first.
High stagnation temperature. ET can reach 200 °C+ internally at zero flow, stressing thermal-expansion joints and (in glycol systems) degrading the antifreeze fluid faster than flat-plate would.

Where evacuated tube genuinely wins

In extreme cold — continental or sub-arctic winters with ambient regularly below −10 °C — or any application demanding high delivery temperature (process heat, district heating, late-winter space heating in northern Europe), evacuated tube is the clearly correct choice. Its lower a₁ heat-loss coefficient means the collector can hold 60–80 °C output even at −15 °C ambient and still capture useful net energy — a flat-plate simply can't sustain that thermal gradient against the heat loss to ambient. When the operating ΔT exceeds the flat-plate/ET crossover point (typically ~30–35 K), ET's vacuum insulation outweighs every other consideration; aperture area, snow handling, and fragility become secondary concerns when nothing else can deliver the temperature you actually need.

For domestic hot water at mid-latitudes (UK, central Europe, most of the continental US), the application doesn't demand those high delivery temperatures, the snow shed-vs-stay-on calculus matters more across the heating season, and the headline-η advantage of ET is largely an aperture-area artefact. Flat-plate is the engineering-correct choice for those climates and use cases.

What the rest of the cascade is

Cosine of incidence (×0.85) — averaged across the solar day, the sun is rarely normal to the collector face; the panel collects the projected component cos(θ). Even with perfect tilt and orientation, half the day's hours are at θ > 45° from normal.
Pump / control losses + morning warm-up (×0.93) — the controller waits until the absorber is warmer than the tank before starting circulation, then the cold pipe loop has to warm up before any useful heat lands at the tank. Both bites come off the morning yield.
Stagnation (×0.95) — for installs sized to deliver useful shoulder-season output, summer at noon often finds the tank already at its hi-limit cut-off and the controller stops pumping. Any further clear-sky energy is simply not collected (drainback systems) or is dumped via a bypass valve (closed glycol systems).
Pipe loss to surroundings (×0.93) — every metre of pipe between collector and tank, especially outdoor runs, leaks heat proportional to (T_pipe − T_ambient) × pipe_UA. Short, well- insulated, indoor runs minimise this.
Soiling, dust, partial shading (×0.97) — accumulates over the year; an annual hose-down recovers most of it.

References & sanity benchmarks

Hottel, H. C., Whillier, A., & Bliss, R. W. (1958). Evaluation of flat-plate solar collector performance. The original heat-balance derivation behind the η = η₀ − a₁·(ΔT/G) form still used on every collector datasheet today.
Klein, S. A., Beckman, W. A., & Duffie, J. A. (1976). A design procedure for solar heating systems, Solar Energy vol. 18, pp. 113–127. The F-Chart paper.
Duffie, J. A., Beckman, W. A., Blair, N. (2020). Solar Engineering of Thermal Processes, Photovoltaics and Wind, 5th ed., Wiley. Ch. 6 (collector performance) and Ch. 21 (F-Chart sizing). Canonical textbook reference.
SRCC OG-100 / Solar Keymark collector certification databases — public source for individual collector η₀ and a₁ values (including the AET AE-32 numbers used in the worked example).
Fraunhofer ISE long-term monitoring of European residential installs: 35–50 % annual delivered efficiency typical.
IEA SHC Task 26 (small DHW systems benchmarking): 38–42 % annual delivered, median across European participants.
Typical per-m² yield benchmark widely reported in Fraunhofer ISE / IEA SHC monitoring: 400–600 kWh/m²/yr of net energy delivered to the store is the band a well-sized mid-latitude flat-plate install lands in. Below ~350 = something is wrong (over-sized array, undersized tank, high delivery temperature, heavy shading). Above ~650 = either the climate is exceptional or the model is being too optimistic.