Why 70% versus 20% is the wrong comparison
Solar thermal converts around 65% of sunlight into heat. Solar PV converts around 20% into electricity. That single statistic drives most of the coverage on this subject, and it is close to meaningless, because the two numbers describe entirely different products.
In 2025 the UK installed 267,032 certified solar PV systems, a third above the record set at the height of the Feed-in Tariff. In the whole of 2024 it installed fewer than 200 solar thermal systems, the lowest figure since records began in 2009. Britain now fits more photovoltaic arrays in a single day than solar collectors in an entire year. This guide explains why, and identifies the narrow set of cases where the market has it wrong.
- The efficiency comparison is invalid. 65% of sunlight turned into low-grade heat and 20% turned into electricity are not comparable outputs. Electricity does anything; heat only heats water.
- Thermal wins per square metre of roof, PV wins per pound of capital. A collector delivers more energy per square metre, but costs roughly five times as much per square metre installed, at £1,418 against £286.
- Real-world thermal output is a third below the brochure. Energy Saving Trust field trials found well-installed systems supplied around 39% of annual hot water on average, against the 50-70% commonly marketed.
- Spend £6,000 on a gas-heated home and the gap is five to one. Solar thermal returns about £146 a year. Solar PV returns about £729.
- Thermal still wins in one situation. Electric hot water plus a small or awkward roof. Below roughly ten usable square metres, a collector beats panels outright.
They do different jobs
Solar photovoltaic panels convert light into electricity. Solar thermal collectors convert light into heat in a fluid. Both sit on a roof, both are dark, both are sold by companies with the word solar in the name, and there the similarity ends.
Electricity is fungible. It runs the kettle, the heat pump, the car and the immersion heater, and anything you cannot use can be sold to the grid. Heat is not. A solar collector produces hot water and nothing else, it cannot be exported, and once the cylinder is full the surplus has nowhere to go.
That difference explains almost every finding below. It is not a story about one technology being better engineered. It is a story about one output being worth more than the other in a modern British home.
- Solar PV
- Photovoltaic panels producing direct current electricity, converted to mains AC by an inverter.
- Solar thermal
- Collectors producing heat in a water and glycol loop, transferred to a hot water cylinder through a coil.
- Solar fraction
- The share of annual hot water demand a collector supplies. The number that matters for thermal, and often lower in practice than in modelling.
- Self-consumption
- The share of PV generation used in the house rather than exported. Batteries, EVs and diverters raise it.
- Smart Export Guarantee
- The scheme obliging suppliers to pay for exported electricity. PV qualifies; solar thermal has no equivalent.
- Diverter
- A device sending surplus PV electricity to the immersion element, letting panels do the collector’s job at low cost.
- Twin-coil cylinder
- A cylinder with two heat exchangers, required for solar thermal and not for PV.
- Stagnation
- Collector overheating when the cylinder is already hot. A thermal failure mode with no PV equivalent.
The efficiency trap
Every article on this subject leads with the efficiency figures, and they are accurate. A good evacuated tube collector converts 50% to 70% of incoming solar radiation into heat. A modern photovoltaic panel converts 20% to 23% into electricity.
The problem is that conversion efficiency measures how well a device does its own job, not how useful that job is. Comparing them is like comparing a kettle’s efficiency to a car’s. A kettle converts nearly 100% of its input into heat and a car converts about 30% into motion, and nobody concludes that kettles are better vehicles.
The right question is not how much of the sunlight each device captures. It is how much money each square metre of roof, and each pound of capital, returns per year. Answer those two and the comparison resolves itself.
Head to head on twelve measures
| Measure | Solar PV | Solar thermal |
|---|---|---|
| Output | Electricity, any use | Heat, hot water only |
| Conversion efficiency | 20-23% | 50-70% |
| Yield per m² per year | ~180 kWh electric | ~300-450 kWh heat |
| Installed cost per m² | ~£286 | ~£1,418 |
| Typical system cost | £6,900-£9,200 | £4,200-£6,500 |
| Return on capital | 10-15% a year | 2-8% a year |
| Revenue side | SEG export income | None |
| Maintenance over 20 years | Close to zero | £800-£1,400 |
| Service life | 25-30 years | 15-25 years |
| Needs a cylinder | No | Yes, twin-coil |
| Works with battery, EV, heat pump | Yes | No |
| UK installations, latest year | 267,032 | Under 200 |
Thermal takes two of the twelve rows, and they are the two that get quoted. The cost per square metre row is the one that decides the outcome, and it is the row nobody publishes.
Per square metre of roof
On a strict energy basis, solar thermal is the better use of a square metre. Four square metres of evacuated tube produces roughly 1,800 kWh of heat a year in modelling. Four square metres of PV, about 0.85 kWp, produces roughly 720 kWh of electricity. Thermal delivers around two and a half times more energy from the same area.
Convert both to money and the ranking depends entirely on what the heat is displacing.
Against an electric immersion, thermal returns nearly three times what PV does from the same roof area. Against mains gas, the two are within a couple of pounds of each other, and a household with a battery or an electric car pulls ahead on PV.
Value per square metre only matters if square metres are the binding constraint. For most British houses they are not: the limit is budget, not roof. That is why the next section, rather than this one, decides the answer for the majority of readers.
Per pound of capital
Here is the number that settles it. A 4 m² solar thermal system at £5,670 works out at £1,418 per square metre installed. A 6 kWp PV array at £8,000 covers about 28 square metres, which is £286 per square metre. Thermal costs roughly five times as much per square metre of roof it occupies.
That gap swamps the energy advantage. Run the same £6,000 into each technology on a mains gas home:
1,800 kWh of heat displacing gas at 8.1p per kWh
4.5 kWp generating 3,825 kWh, half used and half exported
| Scenario | Annual return | Payback | Verdict |
|---|---|---|---|
| PV, battery or EV household | £880 | ~9 yrs | Strong |
| PV, typical gas household | £729 | ~11 yrs | Strong |
| Thermal, electric hot water | £305-£470 | 12-19 yrs | Workable |
| Thermal, LPG | £232 | ~24 yrs | Marginal |
| Thermal, oil | £173 | ~33 yrs | Weak |
| Thermal, mains gas | £95-£146 | 39-60 yrs | No |
The bottom row is the important one, because mains gas heats water in the majority of British homes. On modelled output the payback is 39 years and on measured field-trial output it is nearer 60. Either figure exceeds the life of the collectors.
What thermal actually delivers
Manufacturers and installers typically quote a solar fraction of 50% to 70% of annual hot water. Energy Saving Trust field trials of installed UK systems found that well-installed examples supplied around 39% on average, with the best performers reaching 60%.
That is a substantial gap, and it is not evidence of bad products. It reflects the things modelling struggles with: cylinder standing losses, boilers set to fire before the sun has had a chance, occupancy patterns that draw hot water in the morning rather than the evening, and controllers left on default settings.
The single largest recoverable loss is boiler timing. If the boiler heats the cylinder at 6am, the collector spends the day with nowhere to put its heat. Setting the backup to fire in the late afternoon, after the sun has had its attempt, can move a system several percentage points up the solar fraction range at no cost.
Solar PV has no equivalent gap, because its output is measured at the meter and not inferred from a cylinder. Real-world PV generation lands close to MCS modelling, and where it disappoints, shading is almost always the cause and it is visible on a survey.
The optionality argument
The strongest case for PV is not on any spec sheet. It is that every new electrical load added to the British home over the past five years has increased what a photovoltaic array is worth, and none of them has done anything for a solar collector.
- Batteries. Store surplus generation for the evening peak, lifting self-consumption from around 30% to 70%. A collector cannot charge a battery.
- Electric vehicles. Add roughly 2,100 kWh a year of shiftable demand that panels can serve. A collector cannot charge a car.
- Heat pumps. Turn electricity into space heating and hot water at a coefficient of performance above three, which competes directly with what the collector was for.
- Immersion diverters. For £400 to £600 on top of a PV install, panels do the collector’s job at perhaps 60% of its effectiveness, using roof space you already fitted.
- Export tariffs. Surplus PV earns 5p to 15p per kWh through the Smart Export Guarantee. Surplus heat earns nothing and causes stagnation.
A solar collector installed in 2014 does exactly what it did then. A PV array installed in 2014 is now worth more, because the household around it has acquired new ways to use what it makes. That asymmetry compounds, and it is why the gap between the two technologies has widened rather than narrowed.
What the market decided
UK installation figures are unusually decisive on this question. Solar thermal peaked in the era when the Renewable Heat Incentive paid for it, with monthly volumes regularly hitting a thousand systems in mid-2021. The RHI closed to new applicants in March 2022. Fewer than 900 systems were installed across 2022 and 2023 combined, and 2024 fell below 200, the lowest annual total since records began in 2009.
Over the same period, PV went the other way. 2025 produced 267,032 certified installations, 31% above the record set in 2011 under the Feed-in Tariff, and the first half of 2026 added almost 150,000 more.
The Feed-in Tariff drives PV to 203,125 installations in 2011, while grant support underpins a healthy solar thermal market.
Solar thermal gains a payment per kilowatt-hour of heat generated, and the sector has its strongest years.
PV loses its generation subsidy but keeps a revenue route through export payments. Installation volumes dip, then recover.
Solar thermal loses its only subsidy and gains no replacement. Legacy payments continue to 2032 for existing recipients.
The lowest annual total since MCS records began in 2009, against a cumulative 43,385 systems since then.
267,032 certified PV installations in 2025 and almost 150,000 in the first half of 2026. Neither technology has a capital grant.
Neither technology qualifies for a capital grant in 2026. The Boiler Upgrade Scheme covers heat pumps and biomass boilers only. Both get 0% VAT until 31 March 2027. The difference is that PV has a revenue route through export payments and thermal has none, so the removal of subsidy hurt one far more than the other.
Which to choose, by situation
Scores are out of five and are our editorial judgement for each specific situation, weighted for return on capital, roof constraints and how the household is likely to change over the system’s life.
Solar PV, for the large majority of UK homes
A 4.5 to 6 kWp array with an immersion diverter, which covers hot water alongside everything else the house needs.
The commonest UK situation and the clearest answer. Gas delivers heat at around 8.1p per kilowatt-hour, so displacing it with a collector returns very little, while PV electricity displaces imports at 26.11p and exports the rest. Add a £400 to £600 diverter and the panels handle summer hot water anyway.
Why: Five times the annual return for the same capital
The optionality argument at its strongest. Self-consumption rises towards 70 to 80%, which pushes the value of each generated kilowatt-hour close to the full import price. A collector contributes nothing to either load and competes for the same roof.
Why: Every additional electrical load raises PV’s value and leaves thermal’s unchanged
Thermal returns more per square metre here, but if roof area is not the constraint that does not matter. A larger PV array plus a diverter covers hot water in summer, everything else year-round, and earns export income. Where the roof runs out before the budget does, reconsider.
Why: More total output when area is not limiting
The one domestic case where thermal wins outright. Below roughly ten usable square metres, four square metres of collector returns around £470 a year against about £188 from the same area in panels. Flats with electric cylinders, dormer roofs and heavily shaded properties with one clear patch all qualify.
Why: Nearly three times the return where square metres are the binding constraint
Guest houses, care homes, gyms, campsites and laundries have large steady loads that keep pace with summer output, so stagnation stops being a design limit and collectors run at high utilisation for months. This is where solar thermal remains properly competitive rather than merely defensible.
Why: Demand matches supply across the whole season
A heat pump already produces hot water at roughly 7.5p per kilowatt-hour, so a collector would be displacing heat cheaper than gas. Panels, by contrast, feed the heat pump directly and cut the cost of everything it does. Thermal and heat pumps are alternatives for the same load, not partners.
Why: The collector would be competing with heat that is already cheap
Can you have both?
Yes, and it is rarely the right answer. Three routes exist.
Side-by-side installation puts collectors and panels on different parts of the roof. It works technically and doubles the number of trades, certificates and points of failure, and the collectors are competing for the space that would have gone to panels.
Hybrid PVT panels generate electricity and capture waste heat from behind the cells in one unit. Attractive on a small roof, but cooling the cells means running the thermal loop at low temperature, so you get plenty of tepid water rather than a useful volume of hot water, at £550 to £850 per square metre.
The third route is the one most households should take. A PV array with an immersion diverter does the collector’s job at perhaps 60% of its effectiveness for £400 to £600, uses no additional roof, needs no glycol, no pump and no annual service, and reverts to normal PV behaviour the moment the cylinder is satisfied.
If you already have a working solar thermal system, keep it. The capital is spent and the marginal running cost is a glycol service every five to seven years. This comparison is about where to put new money, not about ripping out functioning equipment.
Pros and cons
- Return on capital, at roughly five to one on a gas home
- Cost per square metre of roof occupied
- Output that can be used, stored, exported or sold
- Near-zero maintenance across 25 to 30 years
- Compatibility with batteries, EVs and heat pumps
- Predictable output that matches the modelling
- Energy captured per square metre, by about 2.5 to 1
- Value per square metre where hot water is electric
- Lower absolute system cost for a small installation
- Summer hot water coverage of 90% or more
- Steady commercial loads that run all year
- Homes where usable roof area is the binding constraint
Work out what a kilowatt-hour of your hot water costs today. Above about 15p, thermal deserves consideration.
Measure your usable roof area before deciding. Under ten square metres changes the answer.
Ask thermal installers for a measured solar fraction, not a modelled one.
Price a PV array with a diverter as the comparison case, not PV alone.
Compare the two on conversion efficiency. It measures different outputs.
Fit thermal to displace mains gas. The payback exceeds the equipment life.
Assume a grant exists for either. Neither qualifies for the Boiler Upgrade Scheme.
Remove a working solar thermal system to fit panels. The capital is already sunk.
Frequently asked questions
Which is better, solar panels or solar thermal?
Solar PV for the large majority of UK homes. Thermal captures more energy per square metre but costs roughly five times as much per square metre installed, so PV returns more for the same money. The exception is a home with electric hot water and under about ten square metres of usable roof, where a collector wins.
Is solar thermal more efficient than solar PV?
Yes, at converting sunlight, and it is the wrong measure. A collector turns 50% to 70% of sunlight into heat while a panel turns 20% to 23% into electricity, but heat only heats water while electricity does anything and can be sold. The useful comparisons are value per square metre of roof and return per pound of capital.
How much does each cost in the UK?
A 4 m² solar thermal system costs £4,200 to £6,500 installed including a cylinder, around £1,418 per square metre of roof. A 6 kWp PV array costs £6,900 to £9,200 and covers about 28 square metres, around £286 per square metre. Both are zero-rated for VAT until 31 March 2027.
What is the payback on each?
PV typically pays back in 9 to 11 years, faster with a battery or an EV. Solar thermal pays back in 12 to 19 years against an electric immersion, around 24 years against LPG, 33 against oil, and 39 to 60 years against mains gas, which exceeds the life of the collectors.
Can solar panels heat my water instead?
Yes, using an immersion diverter costing £400 to £600 fitted alongside a PV array. It sends surplus generation to the cylinder element and does perhaps 60% of what a dedicated collector would, using roof space you have already paid for, with no glycol, no pump and no annual service. For most gas-heated homes this is the better route.
How many solar thermal systems are still being installed in the UK?
Very few. MCS recorded fewer than 200 certified solar thermal installations in 2024, the lowest annual total since records began in 2009, against 43,385 cumulatively. Solar PV recorded 267,032 certified installations in 2025 alone. The divergence dates from March 2022, when the Renewable Heat Incentive closed to new applicants.
Is there a grant for either technology?
No capital grant applies to either in 2026. The Boiler Upgrade Scheme funds heat pumps and biomass boilers only. Both qualify for 0% VAT until 31 March 2027. The practical difference is that PV earns ongoing income through the Smart Export Guarantee while solar thermal has no revenue route at all.
Should I remove my old solar thermal system?
Not if it works. The capital is already spent and ongoing costs amount to a glycol service every five to seven years at £150 to £300. Removing a functioning system to free roof space only makes sense if that space is the constraint on a PV array you actually intend to fit, and even then the arithmetic is usually marginal.
Updated 22 August 2026 with MCS installation data through the first half of 2026, Energy Saving Trust field-trial figures on real-world solar fractions, and the Ofgem July to September 2026 price cap.
The better technology and the worse investment
Solar thermal is the more efficient device. It captures two and a half times more energy per square metre of roof than photovoltaic panels do, and in a home with electric hot water and a small roof it is the right answer. Those are real advantages and they deserve stating plainly rather than being dismissed.
They are also outweighed for most households by two things. A collector costs roughly five times as much per square metre of roof it occupies. And it makes heat, which cannot be stored in a battery, sent to a car, sold to the grid or used for anything except hot water, in a decade when every new load arriving in the British home has been electrical.
Work out the cost of a kilowatt-hour of your hot water, then measure your usable roof. Above 15p and under ten square metres, fit a collector. Anywhere else, fit panels and add a diverter for £500, which does most of the collector’s job on roof space you were using anyway.
METHODOLOGY: Thermal yields modelled at 450 kWh per m² per year for evacuated tube collectors on a south-facing 35-degree roof, with real-world solar fractions taken from published Energy Saving Trust field-trial findings. PV yields modelled at 850 kWh per kWp per year and 0.21 kWp per m². Values use the Ofgem July to September 2026 price cap of 26.11p per kWh for electricity and 7.3p for gas, with 90% boiler efficiency, a 12p export rate and 50% self-consumption unless stated. Installation figures are MCS certified installation counts as reported for 2024, 2025 and the first half of 2026.
DISCLAIMER: Scores are our editorial judgement. Outcomes vary considerably with roof orientation, shading, occupancy, hot water use and tariff; ask installers for an MCS calculation specific to your property before deciding. Prices, VAT treatment and scheme eligibility change; verify the current position before committing. We do not provide financial advice.