Combining different pipe diameters in the heating circuit using reduced fittings
Combining different pipe diameters in the heating circuit using reduced tees
Anyone who gets involved in designing central heating pipe layouts during a renovation or new construction will sooner or later encounter a seemingly trivial but in practice quite tricky question: what to do when pipe diameters must meet at one node and are not the same. This is not an exceptional situation – on the contrary, it is a completely normal part of any well-designed hydraulic system. The main pipe carries a larger volume of heat transfer fluid and naturally has a larger internal diameter, while the branches to individual radiators, zone valves or floor loops are dimensioned for a lower flow and can be – and usually are – one or two sizes smaller.
This is where reduced tees come into play. They are not just "small auxiliary fittings," as some inexperienced plumbers might dismissively call them. They are structural elements that define the hydraulic logic of the entire system, and their correct selection directly affects whether the system will heat evenly, efficiently and quietly, or whether it will be hydraulically unbalanced, noisy and energy inefficient.
In this article, we will take a close look at how to work with different diameters in one circuit in practice, what needs to be considered during planning, which specific diameter combinations occur most frequently and how to work with them during the installation of copper pipe using the Hepworth system.
Why we combine different diameters – basic hydraulic logic
Correct pipe system dimensioning is guided by a simple physical principle: the amount of heat transfer fluid that the pipe must carry determines its internal diameter. If we have a boiler with a power of 20 kW and a temperature drop of 10 °C, the total volume of circulating water is relatively large and the main pipe must be, for example, DN 28. However, if we divide this total power among four radiators, each branch section only needs a quarter of the total flow – and thus can have a much smaller diameter, typically DN 15 or DN 22.
This reduction in diameter after the branches is not just a matter of "saving materials." It has a direct impact on the flow velocity in the pipes, pressure losses, noise and whether the circulating water actually reaches where it is needed. If we were to use the same large diameter everywhere, the flow velocity would drop below the critical limit in the branches, air would be harder to vent from the pipes, and hydraulic balance would become practically unachievable.
Combining different diameters in one system is therefore not a technical compromise – it is the correct engineering practice. And the reduced tee is a fitting that allows this transition elegantly, reliably and in minimal space.
How to read the marking of a reduced tee – what three numbers tell us
Before we move on to specific scenarios, it is good to recall what exactly the three-digit marking of the fitting means. If you have not yet read the article Dimensions of reduced tees: how to read the marking 22 x 22 x 15 from our Knowledge Center, we recommend starting there. Here we will just briefly recap:
- First number – diameter at the inlet (flow coming from the main pipe)
- Second number – diameter at the outlet (continuation of the main pipe further on)
- Third number – diameter of the branch (side arm of the tee)
If the first two numbers are the same, the main pipe only changes diameter in the branch. If the first two numbers also differ, it is a combination of flow reduction and branching – which is less common, but not exceptional. This very detail makes reduced tees very flexible elements, as their correct selection allows us to solve multiple hydraulic situations at once.
Most common diameter combinations in practice – concrete examples
Transition from DN 22 to branch DN 15 – classic when feeding radiators
This is by far the most common situation we encounter during renovations of apartments and houses. The secondary circuit runs in diameter DN 22, while supplying a row of radiators. Each radiator is supplied via a branch DN 15. For this solution, the tee 22 x 22 x 15 is used, where both arms of the main circuit remain in DN 22 and the side branch drops to DN 15.
In practice, this looks like this: for example, a three-bedroom apartment with five radiators, the secondary circuit runs from the distributor in the hallway box. Each radiator has a supply and return branch in DN 15, while the circuit itself between the distributor and the last radiator is run in DN 22. At each radiator branch, we connect one tee 22 x 22 x 15 on the supply and one on the return. The result is a clean, hydraulically correct system, where the DN 15 pipes maintain an appropriate flow velocity (usually in the range of 0.3–0.8 m/s) and noise is minimal.
Branch DN 22 from main pipe DN 28 – feeding zones or larger radiators
A second very common situation arises when a secondary circuit for an entire floor, a group of rooms or a floor loop branches off from the main boiler pipe DN 28. Here we are talking about flows that, although lower than in the main pipe, are still large enough to require a branch in DN 22, not DN 15.
For this case, there is the tee 28 x 28 x 22. The main pipe remains in DN 28 on both sides of the tee, the branch goes into DN 22. Typical application: a family house with two floors, where the ground floor and the upper floor are supplied as two separate zones. From the boiler, the main pipe DN 28 runs, at the first tee a DN 22 branch for the ground floor branches off, the DN 28 pipe continues further and at the second tee a DN 22 branch for the upper floor branches off.
Branch DN 15 directly from main pipe DN 28 – less common, but real
Sometimes it happens that from a larger main pipe we need to branch off directly to a small diameter, for example when feeding a single small radiator in the entrance hall or bathroom from a circuit where it would be uneconomical to introduce an intermediate DN 22. For this case, the tee 28 x 15 x 28 is used – both arms of the main pipe are DN 28, the branch drops directly to DN 15.
It should be noted, however, that with such a combination, it is necessary to carefully check the hydraulic balance. A branch DN 15 from DN 28 has a very low hydraulic resistance compared to other branches, which can cause a preferential flow precisely through this branch at the expense of more distant radiators. In practice, this is solved by a control valve on this branch, which artificially increases the resistance and balances the system.
Reduction in flow and in the branch – variable diameter at both locations
Less common, but occurring in more complex systems, is the situation where the diameter of the main pipe (i.e., the first and second number in the fitting designation are different) and the diameter of the branch change at the same time after a T-piece. Such fittings are used in places where the last branch in the row is located and the pipe behind it serves only one circuit with a lower flow rate.
An example is the T-piece 28 x 22 x 28, where the inlet is DN 28, the main pipe outlet drops to DN 22, and the side branch remains DN 28 – or other combinations according to the design. This is a good example of why it is important to sit with the design and schematic before ordering materials, rather than just estimating by eye.
System sizing before selecting fittings – a proven procedure
Over the years of practice, we have seen dozens of jobs where the plumber came to the site, started laying the pipe and ordered the fittings on the fly. The result was always the same: an ill-considered hydraulic system, the need for corrections, and sometimes a complete rework of the layout. The correct procedure looks different.
Step 1 – heat losses of rooms. Each room must have its heat loss calculated in W or kW. From this, we can determine the power the radiator must deliver and the flow the pipe must ensure to it. Without this number, we cannot size anything.
Step 2 – layout diagram. We draw (at least schematically) the pipe route from the boiler through the manifold to the last radiator. We mark where the branches are, in what order the radiators are fed, and what their mutual distances are.
Step 3 – sizing of sections. For each pipe section, we calculate the flow (the sum of flows of all radiators behind that section) and from that we determine the diameter. A typical approximate flow velocity is 0.3–1.0 m/s for copper. Exceeding 1.0 m/s causes noise, while dropping below 0.2 m/s makes hydraulic balance problematic and air removal difficult.
Step 4 – selection of fittings. Only now, when we know exactly what diameter each section has and where the diameters meet, do we select the specific T-pieces. This is the point where you decide between straight and reduced T-pieces and where the purchase list is finalized.
We write more about the actual selection process in the article How to choose the right reduced T-piece for a heating system and about typical sizing errors in the article Common mistakes in the installation of copper T-pieces and how to avoid them.
Practical scenarios from real jobs
Scenario 1 – Reconstruction of a three-story family house
Job: older family house, 24 kW gas boiler, complete reconstruction of the layout. The original pipe was steel, DN 3/4" and DN 1/2", being replaced by copper. The project specified the following diameters: from the boiler to the manifold DN 28, from the manifold to individual floors DN 22, from floor branches to radiators DN 15.
Solution: Individual circuits for each floor are attached to the manifold. The main pipe from the boiler to the manifold is DN 28. There are no T-pieces at the manifold – each branch runs separately. On the floors, the circuits are in DN 22 with branches to radiators via T-pieces 22 × 22 × 15. In total, we used 12 pieces of T-piece 22 × 22 × 15 on this house (6 on the supply + 6 on the return for 6 radiators). Result: the system works without noise, all radiators heat evenly.
Scenario 2 – New construction with floor heating and radiators
The ground floor has floor heating with multiple loops, and the upper floor has classic panel radiators. 18 kW boiler, combination of circuits. From the mixing unit, DN 22 goes to the floor circuit and DN 22 to the upper floor radiator circuit. Both circuits come from DN 28 boiler circuit.
Solution: From the boiler, the pipe is DN 28, and for the branching to both circuits, T-pieces 28 × 28 × 22 are used. Each T-piece also separates one of the secondary circuits. On the upper floor, there are further T-pieces 22 × 22 × 15 at the radiators. In total, this job was technically cleaner, as the design was done before installation and the fittings were ordered based on the drawings.
Scenario 3 – Simple renovation of an apartment, only radiators
An apartment on the 3rd floor, 4 radiators, 12 kW boiler. Secondary circuit in DN 22, supply and return pipe to each radiator in DN 15. This is the simplest scenario: four T-pieces 22 × 22 × 15 on the supply and four on the return, no larger diameters except the DN 22 supply to the apartment.
The last radiator in the row was interesting – the pipe ends behind it. Instead of a T-piece, a simple straight reducing coupling from DN 22 to DN 15 is used there. This is an important point: a T-piece at the end of the circuit is pointless; a straight reduction or an end piece should be used there instead.
Hydraulic balance of a system with different pipe diameters – what to pay attention to
A system with different pipe diameters is not automatically hydraulically balanced. On the contrary – the presence of reductions creates asymmetries in the system that must be consciously addressed. The most important risk lies in so-called preferred flow: water always finds the path of least resistance, which in practice means that the nearest or largest radiator in the loop will receive too much heat, while the distant or smaller one will remain cold.
A reduced T-piece by itself does not ensure balance – it only allows for proper connection. System balance is ensured by control valves, balancing valves, and properly dimensioned pipe layout geometry. Therefore, it is important that the dimensioning and selection of fittings are done simultaneously with the design of the control system, not separately.
Another important point: every diameter reduction (reduction) creates a local hydraulic resistance. During design, these resistances must be included in the total pressure loss of the loop, so that the circulation pump is properly dimensioned. A practical problem arises when pressure losses in fittings are not considered in the design and the pump turns out to be underdimensioned – the system may work, but "barely operable," water is poorly distributed and some radiators never reach the desired temperature.
Soldering of copper reduced T-pieces – specifics with different diameters
The soldering procedure is described in detail in the article Installation of a reduced T-piece Hepworth: soldering step by step, but when combining different diameters in one fitting, there are some specifics worth highlighting.
When soldering a T-piece where large and small diameters meet (e.g., 28 × 28 × 15), you must be careful with the amount of heat. The wall thickness of the DN 28 pipe is greater and requires longer preheating and more heat. On the contrary, the DN 15 branch heats up much faster. If you apply the same flame intensity to both pipes at the same time, you risk burning or over-soldering the DN 15 connection, while the DN 28 connection is still not hot enough for proper solder flow.
Good technique: preheat both larger arms of the T-piece first, then move the flame to the branch. Always add the solder to the side opposite the flame – if the solder starts to flow from the flame away toward the cooler area, it means the temperature is correct. At the same time, the solder must not burn or create brown or black coatings – this is a sign of overheating.
It is also important to follow the soldering order: if you are soldering the entire T-piece node with adjacent pipes, start with the larger diameter. Heat will distribute better and you will not risk the smaller connection cooling down before you reach it.
Material and quality of fittings – why brand selection matters
Copper fittings from Hepworth belong to products with a long history in the European market. They are made from electrolytic copper according to the EN 1254-1 standard, which ensures precise bore dimensions and consistent wall thickness. In practice, this means that a Hepworth fitting from serial production always "fits" on a properly cut and reamed pipe without play, but also without excessive tightness.
Cheap fittings from unknown manufacturers that appear on the market at much lower prices have one fundamental problem: the dimensional tolerances of the collars are larger. The pipe fits into the fitting with a small clearance, which can result in an uneven flow of solder – the joint looks good from above, but micro-capillaries remain without solder underneath, which leads to slow corrosion and eventually a leak after a few years of operation.
On central heating systems, where the piping runs behind plaster, under the floor, or in wall cavities, the long-term reliability of the joints is crucial. Repairing a leak in such locations is several times more expensive than the price difference between a quality and a cheap fitting. Therefore, we recommend staying with proven products even if the price is slightly higher.
Direction marking of flow – T-piece in flow direction vs. branch position
A reduced T-piece does not have a fixed direction of flow from a physical perspective – it can serve as both a distributor and a collector. In the supply line of hot water, it functions as a distributor (input from the boiler, output to the next radiator + branch to this radiator). On the return line, the same T-piece functions as a collector (water returns from two paths and continues further to the pump and boiler).
This is important practically when you decide which end to mount the T-piece. Hepworth fittings are symmetrical in the direction of the main flow, but asymmetrical between the main flow and the branch. Therefore, you always choose the mounting direction so that the branch arm points towards the radiator (or zone), not by "turning" the fitting upside down and being surprised that it fits geometrically, but is hydraulically unsuitable.
When working with fittings, we recommend visually inspecting them before soldering: whether the inside of the collar is not contaminated, whether the edge is not damaged during transport, and whether the oxide layer is not too thick. With good fittings, these problems do not occur, but it is better to check than to solder blindly.
Combination of reduced T-pieces with other fittings in the system
Reduced T-pieces never appear alone in practice – they are part of a system where they are combined with straight couplings, elbows, straight reductions, compensators, and other components. A few practical notes on this combination:
- Reduced T-piece + straight reduction: If you need to change the diameter again after the T-piece (e.g., from DN 22 to DN 15 after several tens of centimeters), it is sometimes more logical to use a straight T-piece on the larger diameter and then a straight reduction, rather than looking for a T-piece with very specific diameters. It depends on the availability of specific fittings.
- T-piece and elbow: When routing pipe around a corner immediately after a branch, you need to pay attention to the minimum distances between fittings. Attaching an elbow directly after a T-piece is technically possible, but hydraulically it can cause turbulent flow in a critical point. The recommended minimum distance between fittings is at least 5–10 × the pipe diameter in a straight section.
- T-piece and manometer / air vent: Sometimes a reduced T-piece is also used to connect measuring or air venting accessories. In this case, a DN 15 branch is ideal for most standard air vents with an external thread of 1/2".
Typical errors when combining diameters and their consequences
From practice, we know of several recurring errors. We have already mentioned some of them in the article Common mistakes when installing copper T-pieces and how to avoid them, here we add the context of combining diameters:
- Mixing diameters in a T-piece: The installer takes a T-piece 22 × 15 × 22 instead of 22 × 22 × 15, which in practice means that the reduced branch points to the wrong arm. Technically, the pipe fits, but hydraulically it is different. Always check the marking on the fitting.
- Ignoring reduction in overdimensioned piping: The installer uses the same DN 22 diameter throughout the entire route, including all branches to radiators. Result: low flow velocity in the branches, problems with air venting, uneven heating.
- Too large a branch diameter: Using a T-piece 28 × 28 × 22 instead of 28 × 28 × 15 when feeding a small radiator. Result: too high flow through this radiator, other radiators are hydraulically "robbed".
- Wrong soldering order: Starting from the DN 15 branch, both large DN 28 joints cool down and when soldering them, the T-piece is too cold – the solder does not spread properly.
Practical table of diameter combinations and recommended fittings
| Situation in the system | Main pipe | Branch | Recommended fitting |
|---|---|---|---|
| Feeding one radiator from a DN 22 loop | DN 22 → DN 22 | DN 15 | T-piece 22 × 22 × 15 |
| Branching a secondary loop from the boiler pipe | DN 28 → DN 28 | DN 22 | T-piece 28 × 28 × 22 |
| Direct connection of a small radiator from the main DN 28 | DN 28 → DN 28 | DN 15 | T-piece 28 × 15 × 28 |
| Feeding a DN 22 loop with a branch from a larger route | DN 28 → DN 22 | DN 28 | T-piece 28 × 22 × 28 |
| Feeding a radiator from a DN 22 loop with a lateral reduction | DN 22 → DN 22 | DN 15 | T-piece 22 × 15 × 22 |
Material savings and installation time – economic perspective
Reduced T-pieces are slightly more expensive than straight T-pieces because their production is more complex. At the same time, they save time and material by replacing the combination of a straight T-piece and a separate straight reduction. Where we would need two fittings, we can manage with one. This is reflected not only in the material cost, but also in the installation time (one soldering joint instead of two), in the number of potential leak points, and in the overall clarity of the installation.
In a system with 10–15 radiator branches (a typical case for a family house), the use of reduced T-pieces instead of combinations of straight T-pieces and reductions can save 30–40 minutes of installation time and reduce the number of soldered joints by 10–15. At the rate of an experienced installer, this is not an insignificant amount.
Most frequently asked questions (FAQ)
Can I use a reduced T-piece in the reversed orientation – i.e., the branch as the input and the main pipe as the outputs?
Yes, it is physically and hydraulically possible. The fitting works the same regardless of the flow direction – it is just a mechanical connection. Reversed mounting (branch as input) is standard, for example, when collecting return lines. What is important is that the diameters correspond to the actual flow rates – a larger diameter must correspond to a higher flow rate.
What is the difference between the T-piece 22 × 22 × 15 and the T-piece 22 × 15 × 22 – are these two different fittings?
Yes, these are two different fittings. In the first case (22 × 22 × 15), both arms of the main pipe are DN 22 and the branch is DN 15. In the second case (22 × 15 × 22), the inlet arm is DN 22, the outlet arm is DN 15 and the side branch is DN 22 – meaning the reduction occurs directly in the pipe run and the branch has a larger diameter. The geometry is different, as is the installation. Always check the marking before purchase.
Can a reduced T-piece be used in a system with pressure above 6 bar?
Copper fittings from Hepworth are standardly dimensioned for operating pressures up to 10 bar at temperatures up to 110 °C, which significantly exceeds typical conditions in domestic heating systems (usually 1.5–3 bar, temperature up to 90 °C). For special applications (industry, steam), it is necessary to consult the manufacturer's technical data sheet.
What is the minimum distance between two adjacent T-pieces on the same pipe?
There is no fixed standard for domestic installations, but the recommended practice is at least five times the pipe diameter in a straight section between fittings. For DN 22, this means at least 110 mm, for DN 28 at least 140 mm. Shorter distances cause turbulence, increased pressure losses and make soldering more difficult (heat from one joint affects the other).
What if I need a diameter that standard Hepworth T-pieces do not cover – for example DN 35 or DN 42?
For larger diameters (DN 35, DN 42, DN 54), there are separate ranges of fittings. Within the Hepworth system, larger reduced T-pieces are also available – but they need to be searched for separately, as their range is not as extensive as for standard DN 15–28. In some cases, it may be more practical to use a straight T-piece of a larger diameter combined with a separate reduction.
Do I always need to calibrate both pipe diameters before inserting them into a reduced T-piece?
Yes, calibration and deburring (removal of burrs) are mandatory for both or all three ends that you insert into the fitting. Although the fitting can tolerate minor inaccuracies, an uncalibrated pipe may sit at an angle, the joint will have an uneven gap and the solder will not reach it evenly. The result is a weak joint that may look fine from the outside, but is not reliable in the long run.
Conclusion – reduced T-pieces as the foundation of a properly dimensioned system
Combining different diameters in a heating system is not a technical complication – it is an essential part of every properly designed hydraulic system. Hepworth reduced T-pieces are key components in this context, enabling clean, reliable and hydraulically correct connections between pipe sections of different cross-sections. Their selection must always be based on the project, calculated flows and a clear understanding of what diameter each pipe section should have.
A properly selected and correctly installed reduced T-piece is an investment in the long-term reliability and efficiency of the entire system. On the other hand, an incorrectly selected fitting – even if it physically fits – can be the cause of hydraulic imbalance, noise or energy inefficiency, which may only become apparent after years of operation. That is why it is always better to spend enough time on dimensioning before installation, rather than dealing with problems afterwards.
The full range of Hepworth reduced T-pieces can be found in the reduced T-pieces category, where all common diameter combinations for domestic and commercial heating systems are available.
Do you have a question on this topic?
Not sure or dealing with a specific situation in your home? Write to us – we are happy to help.