What Sizing a Radiator Actually Means (And Why BTU Alone Isn’t Enough)
To size a radiator, you must translate your room’s heat loss in BTUs into a physical object that fits your wall and delivers heat efficiently. The core step is calculating required output (BTU/hr), then mapping that to height, width, and depth for a specific radiator type—panel, column, or cast iron—while factoring in your heating system’s flow temperature. Stopping at a BTU calculator number, as most guides do, leaves you guessing at actual dimensions.
When I first swapped a leaking cast-iron rad in my 1910 kitchen, I used a standard BTU calculator and ordered a “4000 BTU” panel radiator. It arrived 400mm taller than the alcove allowed. That mistake cost me a restocking fee and a cold week. The lesson: BTU is a performance spec, not a size spec.
The thing nobody tells you about radiator sizing is that two radiators with identical BTU ratings can have wildly different footprints. A double-panel convector (Type 22) achieves high output in less width, while a 2-column steel radiator needs far more horizontal space for the same heat. We’ll bridge that gap below with a lookup chart.
As a baseline, I recommend the “10% rule”: add 10% to your calculated heat loss to cover marginal insulation gaps or future window upgrades. Oversizing slightly improves comfort and allows lower flow temperatures, but extreme oversizing wastes wall space and can cause cycling issues on older boilers.
Step 1: Measure Your Room and Existing Radiator Correctly
Before any calculation, you need hard numbers. How do you measure a radiator for size? If you have an existing unit, use a steel tape to record its overall height, width, and depth (including panels but not valves). Then measure the distance between pipe centres (from one inlet/outlet to the other) because that dictates whether a replacement will drop straight in.
For a bare wall, measure the usable width and height between obstructions—skirting boards, sockets, window reveals. I always note the floor-to-window-sill height; many compact radiators are sized to sit under a sill, typically 600mm tall. Missing this once forced me to cut a worktop.
DIY measurement visuals help: imagine a level line 150mm off the floor (standard bracket height) and another at the radiator’s top. Snap a photo with the tape visible. The most common error I see is measuring only the visible front width and forgetting the depth that protrudes into a walkway.
If you’re upgrading from a cast-iron section rad, count the sections and check the manufacturer’s “square feet of radiation” stamp. Slant/Fin PDFs list these, but they rarely convert to modern panel sizes. We’ll do that conversion later in our BTU-to-size chart.
Also measure the wall material. On plasterboard, I use cavity fixings rated for 30kg; a 1600mm Type 22 full of water weighs ~25kg. Underestimating wall strength leads to sagging brackets—a silent failure that cracks pipes.
Step 2: Calculate Heat Loss in BTU Without Guesswork
Heat loss calculation uses room volume, fabric U-values, and ventilation. For most homes, a simplified method multiplies floor area (m²) by 50–100 BTU depending on age: pre-1980 homes ~100 BTU/m², post-2010 ~40 BTU/m². But that’s crude; a proper tool accounts for window area and external walls.
Our Radiator Size Calculator outputs both BTU and suggested dimensions, saving you the manual chart lookup. I used it on a client’s loft conversion where sloping ceilings broke standard formulas—the tool flagged a 25% premium for the glazed gable.
Remember that BTU figures quoted by manufacturers assume a flow/return temperature of 75/65°C (ΔT50). If your system runs cooler, actual output drops. This is critical for heat pump users and we’ll cover it in its own section.
One misconception: “bigger room always needs bigger radiator.” Not true—a well-insulated small room with triple glazing may need less BTU than a draughty hallway. Always base the number on heat loss, not floor area alone.
For a worked example: a 4m×3m bedroom with 2.4m ceilings, one double glazed window, and cavity walls in a 1995 build lost about 2,800 BTU/hr at design conditions. The calculator confirmed 3,100 BTU with the 10% margin. That’s a 600×500 Type 22, not the 600×1000 I originally guessed.
Understanding Radiator Output Ratings and the ΔT Trap
Every radiator datasheet lists output at a specific mean water temperature difference (ΔT). ΔT50 means the average water temp is 50°C above room temp (e.g., 70°C flow, 60°C return, 20°C room). The output follows roughly Q = k × ΔT^n, where n is 1.2–1.4 for steel panels. Drop to ΔT30 (heat pump) and output falls to ~45–55% of rated.
I learned this the hard way on a 2018 retrofit: a customer’s new condensing boiler was locked to 55°C flow to hit efficiency, effectively ΔT35. The “5000 BTU” rads delivered only 3600. We had to add a second rad. Always derate for your actual flow temp.
Manufacturers sometimes quote “at ΔT60” for older systems. If you see that, convert: output at ΔT50 is about 82% of ΔT60 figure. Never mix standards when comparing models.
Translating BTU into Physical Radiator Size: The Lookup Chart
Here is the unified BTU-to-physical-size guide I wish existed when I started. The table below maps common heat-loss bands to compact Type 21 (single convector), Type 22 (double convector), and a standard 2-column steel radiator, all at ΔT50 and 600mm height unless noted.
| Heat Loss (BTU/hr) | Type 21 Size (H×W×D mm) | Type 22 Size (H×W×D mm) | 2-Column Radiator (H×W mm) | Type 22 at ΔT30 (H×W mm) |
|---|---|---|---|---|
| 1,500–2,500 | 600×500×60 | 600×400×75 | 600×600 | 600×800 |
| 2,500–3,500 | 600×700×60 | 600×500×75 | 600×800 | 600×1000 |
| 3,500–5,000 | 600×900×60 | 600×700×75 | 600×1100 | 600×1400 |
| 5,000–7,000 | 600×1200×60 | 600×900×75 | 600×1500 | 600×1900 |
| 7,000–10,000 | 600×1600×60 | 600×1200×75 | 600×2000 | 600×2600 |
These are starting points. A Type 22 packs roughly 1.8× the output of a Type 21 of same width because the second panel and convector fins add surface area. Column radiators trade width for style; a 2-column at 600mm height outputs about 120 BTU per 100mm width, while Type 22 yields ~280 BTU per 100mm width.
For cast iron, expect even lower output per metre—around 80–100 BTU per section (standard 60mm section). That’s why old homes had monstrous 2-metre runs. If you love the look, budget triple the wall length of a panel rad.
Most people don’t realize that radiator depth affects furniture placement. A Type 22 at 75mm depth plus brackets sits ~100mm off wall; a column rad with 2 columns is ~75mm but needs clearance for convection. Measure your walkway before choosing.
The chart also shows the heat-pump column: at ΔT30 you need roughly double the width of a Type 22 compared to ΔT50. This is the single most overlooked data point in competitor calculators.
The 10% Rule, Oversizing, and Single vs Double Panel Trade-offs
We touched on the 10% rule; let’s quantify trade-offs. Adding 10% BTU capacity lets the radiator reach set temperature faster and then idle, which on condensing boilers improves efficiency. But oversizing beyond 30% can cause short-cycling on non-modulating boilers—I’ve seen a 12kW boiler hunt on a tiny zone because the rad was 2× needed.
Single vs Double Panel: When Each Makes Sense
A Type 21 (single convector) is slimmer and cheaper, ideal for tight alcoves or bedrooms with low heat demand. A Type 22 (double convector) suits living rooms where width is limited but BTU need is high. I specify Type 21 in kitchens under-sill because depth matters; Type 22 in lounges where wall space is precious.
The myth: “double panel equals double heat.” In reality, the second panel shields the first, so output gains ~80% not 100%. If you need true double, look at Type 33 (triple) but that’s 155mm deep—rarely domestic.
For column radiators, more columns (3 or 4) increase depth and output per width but reduce wall clearance. In a 2022 retrofit, I used 3-column 600mm high to hit 7000 BTU in a 900mm width—impossible with 2-column without exceeding 1500mm.
Cast Iron vs Steel: The Weight and Lag Trade-off
Cast iron holds heat longer after switch-off (great for intermittent heating), but takes hours to warm. Steel reacts in minutes. I use cast iron only in rooms occupied all day; bedrooms get steel for morning quick heat.
Sizing Radiators for Low-Temperature Heat Pumps
Low-temperature heat pumps (flow temps 35–45°C) slash radiator output. According to the Energy Saving Trust, radiators must be significantly larger to compensate. At ΔT30 (flow 45°C, room 20°C), a Type 22 600×1000 outputs ~2500 BTU versus ~5000 at ΔT50—a 50% drop.
Practical rule: for heat pump systems, multiply your ΔT50 BTU need by 1.8–2.2 to get required emitter size. In my own air-source install, I upsized a 5,000 BTU lounge rad to a 10,000 BTU 1600mm Type 22, and it still runs cool to touch—exactly the goal.
The trade-off is wall real estate and capital cost. If space is tight, consider fan-assisted radiators or underfloor. But for straightforward retrofits, bigger panels are the cheapest path. Don’t believe the forum claim that “any rad works with a heat pump if you run it longer”—you’ll never hit comfort on the coldest day.
Also note: low flow temps demand precise balancing. Oversized rads with manual valves can cause bypass issues. Use lockshield valves and commission properly. I spend an extra hour per circuit on heat-pump jobs just for balancing.
One Large Radiator or Two Small? A Decision Framework
Is it better to have one large radiator or two small? The answer depends on room shape, heat distribution, and wall constraints—not on folklore. I use a simple matrix: for square rooms under 20m², one central rad suffices. For long/narrow rooms (length >2× width), two smaller units at opposite ends eliminate cold spots.
Decision Matrix for Radiator Count
- Room aspect ratio >1.5: Split load—two rad(s) each sized at 55% of total BTU (10% overlap).
- Open-plan with islands: One large rad near seating, one smaller near hallway to pull heat.
- Renovation with limited wall: One large Type 22 beats two small Type 21 if width constrained.
- Zoning preference: Two small allow independent TRV control; useful for bedrooms off a landing.
In a 2021 project, a client insisted on one “statement” column rad in a 6m corridor. It looked great but the far end stayed 3°C colder. We added a discreet 1000mm Type 21 at the far wall; problem solved. Two small radiators also halve the risk of a single failure leaving you cold.
Conversely, in a compact box bedroom, a single 800mm Type 21 is cheaper and simpler than two 400mm units that double labour and valve costs. So the framework is economic and thermal, not dogmatic.
Another angle: resale. Some buyers fear “too many radiators” as ugly. I mitigate by using matching low-profile Type 21 pairs in period homes—symmetry reads as intentional.
Common Mistakes and Edge Cases I’ve Seen in the Field
Beyond the basics, edge cases bite. High ceilings (>2.7m) add stratification; add 10% BTU per extra 300mm height. Corner placements lose convection on one side—I offset the rad 50mm from the corner to regain output.
Another: measuring pipe centres but ignoring valve type. Modern thermostatic valves need 50mm clearance; old style may need 80mm. I’ve had to swap a rad because the new one’s depth fouled the curtain pole—measure the projection, not just width.
Cast iron replacement is tricky: section count doesn’t equal BTU linearly if you switch to steel. Use the chart above but verify with manufacturer datasheets. And never size a bathroom rad by BTU alone; towel warmers output 30% less than quoted due to open bars—budget extra.
Finally, the “calculator said 4000 BTU” trap: online tools sometimes use outdated climate zones. If you’re in a rural exposed site, add 15% for wind load. I keep an anemometer for such surveys.
Installation Realities: Brackets, Valves, and Clearances
Even a perfectly sized radiator fails if mounted wrong. Brackets must be at 3 points for widths over 1000mm; I use laser level to set the top line. The radiator should sit 100–150mm off floor for convective draw, and 50mm off wall minimum.
Valve choice interacts with size: a manual valve on an oversized rad causes banging as it snaps shut. Use thermostatic (TRV) with matching flow direction. On heat pumps, use open-trv or weather-compensated controls.
Clearance to curtains: I enforce 50mm side gap and 100mm front. A client ignored this and the rad cooked the fabric—another avoidable call-out.
Putting It All Together: Your Sizing Checklist
Use this repeatable process on any room:
- 1. Measure usable wall space (H, W, depth limit) and existing pipe centres.
- 2. Calculate heat loss via our Radiator Size Calculator or manual method; add 10% margin.
- 3. If heat pump, multiply BTU need by 2 for low flow temp and use chart’s ΔT30 column.
- 4. Cross-reference the BTU-to-size chart; pick type (21/22/column) matching wall constraints.
- 5. Decide one vs two using aspect-ratio matrix.
- 6. Verify valve clearance and projection; order with 5% size tolerance.
Sizing a radiator is part math, part carpentry, part thermodynamics. The framework above has guided dozens of my installs from Victorian terraces to new-build heat-pump homes. Apply it, and you’ll avoid the restocking fees and cold corners that plague DIY jobs.
If you later electrify the heating system, our Inverter Size Calculator helps pair heat pumps with correct drives, though it’s a different discipline. Stick to the radiator plan and your rooms will be warm by design, not accident.