Usually yes with 10 mm copper, rarely with 8 mm. A heat pump moves three to four times the water a boiler does at the same output, because it runs a 5 °C temperature difference instead of 20 °C. At 1 m/s and ΔT 5 °C, 10 mm copper carries about 1.3 kW and 8 mm about 0.75 kW — enough for individual radiators, not for a whole circuit.
Why does microbore matter more with a heat pump?
Because of flow rate, not pressure. The heat a circuit can deliver is set by the flow triangle: kilowatts equal flow rate multiplied by temperature difference multiplied by the specific heat of water. A gas boiler runs a 20 °C difference between flow and return; a heat pump typically runs 5 °C, because a wide ΔT would force the flow temperature up and the COP down.
Same heat, quarter of the temperature difference, four times the water. Pipework that comfortably fed a radiator from a boiler may be running at three or four times its original velocity on a heat pump — and velocity is what causes noise, erosion and unmanageable pressure drop.
How much heat can 8 mm and 10 mm actually carry?
Keep velocity between 0.5 and 1.5 m/s, with 0.9–1.0 m/s as best practice. At those limits:
| Copper pipe (bore) | Flow at 1.0 m/s | Heat at ΔT 5 °C | Heat at ΔT 8 °C |
|---|---|---|---|
| 8 mm (6.8 mm) | 2.2 l/min | 0.76 kW | 1.2 kW |
| 10 mm (8.8 mm) | 3.6 l/min | 1.3 kW | 2.0 kW |
| 15 mm (13.6 mm) | 8.7 l/min | 3.0 kW | 4.9 kW |
| 22 mm (20.2 mm) | 19.2 l/min | 6.7 kW | 10.7 kW |
Read that alongside the whole-system flow: at ΔT 5 °C, a 5 kW heat pump needs roughly 14 l/min, a 10 kW unit about 29 l/min. Those figures cannot pass through 10 mm pipe at any acceptable velocity, which is why the distribution pipework — the run from the heat pump to the manifold or to the first tee — always has to be 22 mm or 28 mm even in a microbore house.
The distinction that decides most projects is therefore this: 10 mm final connections to individual radiators are usually fine; 10 mm as the main circuit is not. A 1.3 kW limit covers a bedroom or a bathroom radiator comfortably and covers nothing in a lounge.
What about plastic microbore, and pipe buried in screed?
Plastic is worse than copper at the same nominal size. A 10 mm plastic pipe has thicker walls and a bore nearer 7 mm, cutting capacity by around a third against 10 mm copper, and push-fit inserts narrow it further at every joint. Treat 10 mm plastic as 8 mm copper for design purposes.
Buried and inaccessible pipework changes the economics rather than the physics. Where microbore disappears into a solid floor, the practical options are to keep it for the small rooms it can serve, to run new pipework in a different route, or to accept a higher ΔT and a higher flow temperature on that branch only. Long runs are the real killer: pressure drop rises with the square of velocity and linearly with length, so 15 m of 8 mm behaves very differently from 3 m of it.
What are the workarounds?
- Zone the problem. Serve small rooms through the surviving microbore and re-pipe only the two or three high-output radiators. This is usually the cheapest competent answer.
- Raise ΔT selectively. Designing a branch at 8 °C rather than 5 °C cuts its flow requirement by about 40%. It costs a little efficiency because the mean radiator temperature falls, so it is a targeted tool, not a global setting.
- Increase emitter surface instead of flow. A larger radiator delivers the same heat at a lower mean water temperature and therefore the same flow — see do I need bigger radiators for a heat pump?
- Convert from single-pipe or radial microbore to a two-pipe layout where the manifold arrangement makes it easy; radial microbore from a central manifold is often simpler to re-pipe than it looks.
- Do not simply raise the pump speed. Pushing 8 mm pipe to 2 m/s buys you noise, erosion of the copper and a circulation pump consuming enough electricity to dent the seasonal COP.
How do I check my own house?
Find the pipe entering a radiator and measure the outside diameter: 8 mm and 10 mm are unmistakably thinner than the 15 mm standard. Then check where it comes from — radial microbore usually runs from a manifold in an airing cupboard or under the floor, while 15 mm distribution with 10 mm tails is a far easier retrofit than 8 mm throughout.
It is worth saying plainly from the manufacturing side: a heat pump sized correctly in kilowatts can still fail if nobody checked the litres per minute, which is why Solimpeks states the water circulation rate alongside capacity for its heat pump range.
Give those findings to whoever produces the heat load calculation. A competent design will state the design flow rate, the design ΔT and the velocity in each branch, and will identify which pipes must change before the quote is priced rather than after the heat pump is installed. That level of detail is also the fastest way to tell two quotes apart; see how do I compare two heat pump quotes fairly?
Frequently asked questions
Do I have to replace microbore pipes for a heat pump?
Not all of them. The distribution pipework must be 22 mm or 28 mm to carry the whole flow, but 10 mm copper tails to individual radiators up to about 1.3 kW usually stay. Only 8 mm and 10 mm plastic frequently need replacing outright.
How much heat can 10 mm pipe carry with a heat pump?
About 1.3 kW at 1 m/s with a 5 °C temperature difference, or roughly 2 kW if the branch is designed at 8 °C. That covers most bedroom and bathroom radiators but not a living room emitter.
Why does a heat pump need so much more water flow than a boiler?
Because it runs a 5 °C difference between flow and return instead of a boiler's 20 °C. Delivering the same kilowatts with a quarter of the temperature difference requires roughly four times the flow rate through the same pipes.
Can I just turn the circulation pump up to push water through microbore?
No. Above about 1.5 m/s copper becomes noisy and subject to erosion corrosion, and the pump's own electricity consumption starts eroding the seasonal COP. Reducing flow requirement through larger emitters or a higher branch ΔT is the correct fix.