How to correctly choose a branch reducer when supplying a radiator from the main pipe

We are still preparing products.

But you can look at other categories.

How to correctly choose a branch reducer when connecting a radiator to the main pipe

When heating circuits are distributed from the main collector or riser pipe to individual radiators in a family house or apartment building, one of the most important — and at the same time most frequently underestimated — decisions is the selection of the correct branch diameter. The seemingly simple question "which T-piece should I use here?" hides a whole range of hydraulic, sizing, and installation-related considerations that directly affect whether the radiator will be adequately supplied with heat or the entire circuit will suffer from imbalance, noise, and insufficient performance.

In this article, we will look at the entire issue of choosing a branch reducer systematically — from the basic understanding of what a reduced T-piece actually does, through hydraulic criteria, typical dimensions and combinations in practice, to concrete examples from field installation and tips on where plumbers most commonly make mistakes.

What is a reduced T-piece and why a straight one is not enough

A straight (non-reduced) T-piece has all three ports of the same diameter. A reduced T-piece has at least one port with a different diameter than the other two. In heating practice, this most often means that the main axis (pass-through) remains larger — for example, 22 mm — and the branch is reduced to 15 mm, because the radiator requires less flow than the main branch supplying other radiators downstream.

If we used a straight T-piece 22 × 22 × 22 at each radiator connection from the 22 mm main distribution and then added a separate reduction from 22 to 15 mm before the radiator, the result would be functional, but unnecessarily complicated: more connections, more space, more soldered material, and a higher probability of leakage. A reduced T-piece combines these two functions (branching + reduction) into one fitting. For a more detailed discussion of the difference between straight and reduced T-pieces in copper piping, see the separate article Difference between straight and reduced T-pieces in copper piping in this Knowledge Center.

Key principle: the branch diameter must correspond to the actual flow

The basic hydraulic rule is simple: the higher the power the radiator requires, the greater the flow of the heat transfer medium it needs, and the larger the diameter of the supply (and at the same time return) pipe to it we must choose. The radiator's power depends on its surface area, number of sections, type — panel, finned, or cast iron — and on the system's operating temperature difference (typically 75/65°C, 70/55°C, or nowadays more common 55/45°C for heat pumps).

In practice, in low-temperature systems (heat pump, condensing boiler in an equithermal mode), we encounter a lower temperature difference, which means that to achieve the same power, we need to circulate more water — and thus we need a larger diameter. This factor is very important in new installations in recent years: what worked with a 15 mm branch at 75°C twenty years ago may not be sufficient even for a smaller radiator at 45°C supply temperature today.

Approximate flow values and corresponding diameters

For basic orientation, the following values apply (at a temperature difference of 10 K and a flow velocity of up to 0.5 m/s, which is the recommended limit for copper in heating from the perspective of noise and erosion):

Copper pipe 15 mm (internal diameter approx. 13 mm): maximum recommended flow approx. 200–250 l/h → suitable for radiators up to approx. 1.5–2 kW at a temperature difference of 10 K
Copper pipe 22 mm (internal diameter approx. 19.6 mm): maximum recommended flow approx. 500–700 l/h → suitable for larger radiators or sections supplying multiple radiators with a total power up to 5–6 kW
Copper pipe 28 mm (internal diameter approx. 25.6 mm): maximum recommended flow approx. 1,000–1,400 l/h → main distribution lines, risers, supply of larger branches

From these values, a direct answer to the question of the branch diameter follows: if you are supplying one radiator up to 1.5 kW, a 15 mm branch is sufficient. If the radiator is larger, or the branch supplies a group of radiators, considerations about a 22 mm branch are appropriate.

How to read the designation of a T-piece and what the numbers mean

Before moving on to selecting a specific size, it is important to understand the notation. A reduced T-piece is always marked with three numbers in the format A × B × C, where:

A = diameter of the first end of the main axis (inlet, supply side)
B = diameter of the second end of the main axis (outlet, side of the next radiator or end of the supply line)
C = diameter of the branch (perpendicular branch, towards the radiator)

For example, the T-piece 22 × 22 × 15 has both ends of the main axis at 22 mm and a branch of 15 mm — this is by far the most commonly used type in family homes, where the main supply is 22 mm and each radiator is supplied via 15 mm. A more detailed explanation of this numbering system can be found in the article Dimensions of reduced T-pieces: how to read the designation 22 × 22 × 15.

With asymmetric T-pieces, where the reduction is also on the main axis (not just on the branch), we encounter notations such as T-piece 22 × 15 × 22 — here, the central neck is narrowed, which is less common but used in specific configurations, for example, when reducing the flow through a section with the same branch. More on this is covered in the section below.

Typical practical scenarios: which T-piece for which situation

Scenario 1: Family home, 22 mm main supply, radiators supplied with 15 mm

This is the absolute most common case in Central European installation practice. A boiler with a power of 15–24 kW, a supply loop made of copper 22 mm, individual radiators with a power of 0.5–1.5 kW supplied via Cu 15 mm. At each branch from the main 22 mm supply to a radiator, it is a T-piece 22 × 22 × 15. The main axis remains 22 mm, the branch is 15 mm — the flow to the radiator is typically 40–120 l/h, which a 15 mm pipe can handle without problems.

In practice, I have seen houses where the installer used only this one type of T-piece throughout the entire house — and the system worked properly, because the sizing was consistent. A bathroom radiator of 0.8 kW, a living room radiator of 1.2 kW, a bedroom radiator of 0.6 kW — all supplied from a 22 mm main supply via 15 mm branches.

Scenario 2: A large panel in the living room or in an open kitchen-living room

With panel radiators of type 22 VK or 33 VK with a power of 2–4 kW, a 15 mm branch is at or beyond the recommended values, especially at lower heating water temperatures (heat pump). In this case, the correct choice is a 22 mm branch, while the main supply can be 28 mm. The appropriate T-piece is T-piece 28 × 28 × 22 — main axis 28 mm, branch 22 mm.

I came across a case where the customer had a living room with a large 33 VK panel (power of about 3.2 kW at 55/45°C) and the previous installer used a 15 mm branch. The radiator never heated evenly — the left side was significantly warmer than the right, because the water simply could not flow fast enough through the narrow 15 mm branch. The customer complained about the "silent" radiator. After replacing the T-piece with a 28 × 28 × 22 and connecting the radiator with 22 mm piping, the problem was resolved.

Scenario 3: Main supply 28 mm, supplying smaller radiators

In older apartment buildings or in boiler rooms supplying multiple floors, we often encounter main risers of 28 mm. When branching off from this riser to a smaller radiator supplied by 15 mm piping (for example, a 0.4 kW bathroom ladder radiator), the correct choice is T-piece 28 × 15 × 28 — it maintains 28 mm on the main axis and narrows the branch directly to 15 mm without the need for further reduction.

Less experienced installers sometimes reach for a 28 × 28 × 22 and then add a reduction from 22 to 15 mm — which is functional, but unnecessarily complicated and adds an extra joint to the system. The fewer the joints, the lower the risk in long-term operation.

Scenario 4: Asymmetric reduction on the main axis — T-piece 22 × 15 × 22

A type such as T-piece 22 × 15 × 22 (narrowed central neck, outer ends 22 mm, branch 22 mm) is a special case. It is used when we want to create a branch with the same diameter as the main branch, while at the same time narrowing the flow in one direction — for example, to simulate increased resistance for hydraulic balancing, or in special configurations of distribution manifolds. In the typical practice of family homes, you will not encounter it very often, but in industrial and administrative buildings you will.

Hydraulic balancing and its relationship to the branch diameter

One of the reasons why some radiators in a house heat properly and others only warm up slightly is incorrect hydraulic balance — that is, a situation where the water seeks the path of least resistance (usually those closest to the pump or boiler) and distant radiators remain unheated. The diameter of the branch is one of the factors forming the hydraulic resistance of the branch.

A smaller branch diameter = higher resistance = less water will flow into that radiator. This can sometimes be intentional (when we want to "slow down" a radiator that would otherwise be overheated and take flow away from distant radiators), but in most cases, we should use balancing valves for balancing, not just the choice of T-piece diameter. If the T-piece is too small, we become dependent on this "accidental" limitation, which, with changes in operating pressure or pump performance, causes unpredictable system behavior.

The correct procedure is: select the branch diameter according to the actual required flow for the given radiator, and solve hydraulic balancing with control elements (thermostatic heads, pre-set balancing valves). More about combining different diameters in the heating circuit can be read in the article Combining different diameters in the heating circuit using reduced fittings.

Material and soldering: why the quality of the connection also matters

Hepworth copper T-fittings are made from CW024A (equivalent Cu-DHP) copper, which ensures good solderability with both hard and soft solder. In heating systems, soft solder (tin-based, typically S-Sn97Cu3 or similar) with a soldering temperature of 230–250°C and active flux is used. In reduced T-fittings, it is important to note that different pipe diameters can cause different heating rates during soldering — the thinner wall of the 15 mm pipe heats up faster than the 28 mm pipe.

In practice, this means: when soldering a 28 × 22 × 28 T-fitting, start heating on the larger pipe and only after the fitting is sufficiently warm, proceed to the smaller one. Otherwise, the smaller pipe may overheat the solder before it flows into the capillary gap of the larger pipe. A detailed step-by-step soldering procedure is described in the article Installation of a reduced Hepworth T-fitting: step-by-step soldering procedure.

When to choose a branch with the same diameter as the main pipe

There are situations where it is not suitable to reduce the branch to a smaller diameter, but rather to keep it the same as the main pipe. This occurs when:

The branch supplies not only one radiator, but an entire secondary branch (e.g., an entire floor or wing of a building) with multiple radiators — in this case, the secondary branch needs the same capacity as the main pipe.
The radiator has a very high output and requires maximum flow — typically a large bathroom panel radiator with a power output over 2 kW or an industrial plate radiator.
The system operates with a low temperature difference (heat pump, underfloor heating in combination with radiators) and the flow must be increased to achieve the required thermal output.

In such cases, we use an equal T-fitting (not a reduced one), or a variant where the branch is only one size smaller than the main pipe — for example, main pipe 28 mm, branch 22 mm (fittings 28 × 28 × 22). Another interesting configuration is the 28 × 22 × 28 T-fitting, where only part of the main pipe is narrowed — this is suitable for series circuits with intentional throttling of flow in a certain section.

Most common mistakes when selecting the branch diameter

From installation practice, several recurring mistakes can be identified that lead to problems in operation:

Undersizing the branch for large radiators: The installer uses a 15 mm branch "because that's how it's always done" — regardless of the actual radiator output. Result: a cold radiator, noise from flow, and an unsatisfied customer.
Overdimensioning the branch for small radiators: On the other hand, some installers use a 22 mm branch even where a 15 mm would suffice — "better bigger". Result: a radiator near the boiler "takes" the flow and distant radiators suffer. A larger branch diameter = lower hydraulic resistance = higher flow through this radiator, which can disrupt hydraulic balance.
Mixing up the direction of the T-fitting: A reduced T-fitting with a narrowing on one end of the main pipe (e.g., 22 × 15 × 22) may be installed in the wrong direction — which changes the hydraulic properties of the connection. Always make sure which direction is the "inlet" and which is the "outlet" from the main line. Typical installation errors are discussed in detail in the article Common installation errors with copper T-fittings and how to avoid them.
Ignoring the system temperature difference: A branch dimension calculated for a system of 75/65°C is insufficient for a system of 45/35°C with the same radiator output. When modernizing a boiler room (e.g., switching from a gas boiler to a heat pump), pipe dimensions must be re-evaluated.
Combining different materials without transition fittings: Copper and steel must not be connected directly in the direction of flow, with copper always coming after steel (from the flow perspective), not before, due to galvanic corrosion. When combining systems, use the correct transition fittings.

Practical step-by-step procedure for selecting a T-fitting

To select the correct reduced T-fitting for each radiator connection, follow this verified procedure:

Determine the radiator output — from the technical data sheet or calculated by the designer. Note: the output stated by the manufacturer is for Δt 50 K (i.e., 75/65°C); for lower temperatures, you must adjust the output according to the correction factor.
Determine the system temperature difference — 75/65°C, 70/55°C, 55/45°C or another. This determines the flow you need to ensure.
Calculate the required flow — formula: Q [l/h] = P [W] / (1.163 × Δt [K]), where Δt is the difference between supply and return temperature.
Compare with pipe capacities — according to the table in the section above, determine which branch diameter is needed.
Check the main pipe diameter — the branch must not be larger than the main pipe (this physically does not make sense for a reduced T-fitting, but it is something to be aware of).
Select the type of T-fitting — according to the combination of the main pipe diameter and the required branch, find the appropriate T-fitting in the category Reduced T-fittings.

Examples of calculation for common situations

Example 1: Bathroom panel radiator, output 800 W at 75/65°C. The system operates at 65/55°C (Δt = 10 K). Flow = 800 / (1.163 × 10) = 68.8 l/h. The main pipe is 22 mm (capacity up to 600 l/h). Required branch: 15 mm (capacity up to 250 l/h, which is sufficient). Choice: 22 × 22 × 15 T-fitting.

Example 2: Large living room panel radiator 33 VK, output 2 500 W at 75/65°C. Heat pump system operates at 50/40°C (Δt = 10 K). The actual required output at the lower temperature is higher — the correction factor for 50/40°C compared to 75/65°C is approx. 0.44, so the radiator must have a nominal output of at least 2500/0.44 ≈ 5 680 W at 75/65°C, in order to deliver 2 500 W at 50/40°C. Flow rate at 50/40°C = 2500 / (1.163 × 10) = 215 l/h. Main pipe diameter is 28 mm. Required branch: 22 mm (capacity up to 700 l/h). Choice: T-piece 28 × 28 × 22.

Example 3: Small corridor radiator, output 350 W. System 70/55°C (Δt = 15 K). Flow rate = 350 / (1.163 × 15) = 20 l/h. Main pipe 28 mm. Branch: 15 mm is more than sufficient with a large margin. Choice: T-piece 28 × 28 × 15, or if a T-piece is followed by the end of the pipe run, consider T-piece 28 × 15 × 28 — it depends on the direction of the pipe run.

Relationship between branch diameter and noise in the system

One of the less obvious, but well-observed phenomena in practice: an undersized branch can be a source of noise in the heating system. When water flows too quickly through a narrow branch (speed exceeding 1 m/s), turbulent flow occurs, resulting in noise — typically described by customers as "noise behind the radiator" or "knocking in the pipes". Copper is not a good sound insulator, so these sounds are transmitted further into the structure.

The recommended maximum flow velocity in copper heating installation is 0.5 m/s (for continuously supplied sections) up to 1 m/s (for short branch sections). If the flow rate is, for example, 200 l/h and the branch diameter is 15 mm (internal diameter 13 mm, cross-section 1.33 cm²), the velocity = 200/(3600 × 0.0001327) = 0.418 m/s — acceptable. If we were to push the same flow rate through a 10 mm branch (which is not used in standard copper installation, but illustrates the principle), the velocity would be three times higher and the noise would be noticeable.

Special cases: two-pipe vs. one-pipe systems

An important context for selecting a T-piece is the type of heating system connection. In a two-pipe system (today's standard), each radiator has its own supply and return pipe — a reduced T-piece is used for the supply pipe and the same type (or a straight T-piece) for the return pipe. The branch sizing is done separately for each pipe.

In a one-pipe system (older system, still encountered in apartment renovations), the radiator is connected in series within one loop. Here, the T-piece (or a special shaped element) is used differently and the sizing is more complex — the flow through the branch depends on the hydraulic resistance of the radiator and the bypass branch. In such cases, it is recommended to leave the sizing to an experienced designer.

Most frequently asked questions (FAQ)

Can I use a T-piece 22 × 22 × 15 for a radiator with an output of 2 kW in a system operating at 75/65°C?

With a temperature drop of 10 K and an output of 2 kW, the required flow rate is approximately 172 l/h. A 15 mm pipe can handle up to 250 l/h at an acceptable velocity, so it is barely acceptable — but only if the system actually operates at 75/65°C and not lower. At a lower supply water temperature (e.g., 55°C), you would need to achieve the same output with a higher flow rate (approximately 290 l/h at Δt 6 K), which a 15 mm branch cannot handle without increased velocity and noise. For radiators above 1.5 kW in low-temperature systems, we recommend a 22 mm branch.

What happens if I choose an undersized branch — will the radiator at least provide some heat?

Yes, the radiator will provide some heat, but it will not reach its nominal output. Moreover: because less water flows through the small branch, the water cools down more in the radiator (higher temperature drop across the radiator), which can cause uneven temperature distribution across the radiator surface — a typical phenomenon of "cold lower part, warm upper part". In severe underdimensioning, condensation of moisture on the colder part of the radiator can occur during transitional periods.

Should I use a reduced T-piece at every radiator, or is a straight T-piece with a separate reduction sufficient?

A reduced T-piece is always the preferred choice — it reduces the number of joints (and thus potential leak points), saves space in the pipe groove, and simplifies installation. A separate reduction after a straight T-piece is a solution you can use if you don't have the correct reduced T-piece available, but it is a compromise, not an ideal solution. In professional installation, a straight T-piece combined with a separate reduction is not used at radiators if a reduced T-piece is available.

Is a T-piece 28 × 22 × 28 the same as a T-piece 28 × 28 × 22?

No — these are different shaped fittings with different geometries. In the T-piece 28 × 28 × 22, the reduced branch (perpendicular branch) is present, while the main pipe remains 28 mm on both sides. In the T-piece 28 × 22 × 28, the central neck on the main pipe is reduced, while the branch is 28 mm, the same as the end necks. These two types are used for completely different hydraulic situations and are not interchangeable.

Do I need special tools when joining copper T-pieces?

For capillary soldering of copper T-pieces, you need: a pipe cutter (not a saw — the cut must be perpendicular and without burrs), a deburring tool, a wire brush or fine-grit sandpaper (for cleaning the connection surfaces), a propane-butane or acetylene torch, flux and solder. For compression joints (as an alternative to soldering), you need a compression plier appropriate for the given size. For 28 mm and larger diameters, compression is faster and increasingly popular in practice. Details on the installation procedure can be found in the article Installation of a reduced T-piece Hepworth: soldering step by step.

Can I use the same reduced T-piece for the supply and return pipe of a radiator?

Yes, in a two-pipe system, the supply and return pipes are symmetrical in terms of sizing — the same flow rate enters and exits the radiator. Therefore, you will use the same type of T-piece (e.g., 22 × 22 × 15) on both the supply and return pipe. The only difference is the direction of flow in the return pipe, but that does not affect the choice of T-piece.

Conclusion: Choosing the correct branch size is the foundation of a functional heating system

Selecting the correct branch size when connecting a radiator to the main pipe is not just a technical formality — it is a decision that directly affects the performance of the entire heating system, its noise level, hydraulic balance, and long-term reliability. The key is to correctly calculate the required flow rate for each radiator, choose the appropriate branch diameter, and use the corresponding reduced T-piece — whether it is a 22 × 22 × 15 for standard family homes, 28 × 28 × 22 for larger radiators, or 28 × 15 × 28 when connecting a small radiator to a large vertical pipe run.

Investing time in a proper sizing calculation — even if it's just five minutes with a calculator before purchasing the fittings — always pays off: you will save yourself from having to redo the installation, from an unsatisfied customer, and from unnecessary costs. And if you are unsure, it is always better to consult with a specialist or read the relevant articles in this Knowledge Center, for example How to choose the correct reduced T-piece for a heating system or When to use a reduced T-piece instead of a straight one in a heating system.

Do you have a question about this topic?

Struggling to decide or dealing with a specific situation in your home? Write to us — we are happy to help.