The Microbore Heat Pump Diagnostic: Flow Velocity Limits and Hydraulic Volumetric Workarounds

Why Microbore Is the Hydraulic Weak Link in a Heat Pump Retrofit

Microbore pipework, typically 8 mm or 10 mm outside diameter soft copper, was widely installed in British homes from the 1970s through to the 1990s. Its appeal to the installers of the day was practical: it was cheap, flexible, easy to route beneath floorboards and through joists without lifting large sections of structure, and it was perfectly adequate for the hydraulic regime that a gas or oil boiler imposed. A central heating system served by a boiler was designed around a high flow temperature, commonly 70 to 80 degrees Celsius, and a wide temperature differential across the system, conventionally 20 K. Under those conditions, a very small mass flow of water is sufficient to carry a useful thermal output to each radiator, and the narrow pipes were not a limiting factor.

The introduction of a heat pump changes the hydraulic premise entirely. The coefficient of performance of a heat pump falls as the temperature lift between the source and the delivered water increases, so the design objective is to run the lowest practicable flow temperature, and to return the water to the heat pump as cool as the emitters allow. The resulting emitter circuit operates on a narrow differential, commonly 5 K and sometimes up to 7 or 8 K, with flow temperatures in the range of 35 to 50 degrees Celsius. This audit characterises the physics linking that narrow differential to the required flow rate, quantifies the resulting fluid velocities in microbore, explains the erosion-corrosion and pressure-drop consequences, and sets out the engineering architecture of hydraulic separation that allows a high-volume primary heat pump circuit to coexist with a restricted microbore secondary network.

The Volumetric Flow and Delta T Calculus

The thermal power carried by a flowing liquid is governed by a single energy balance: Q = m-dot × c × ΔT, in which Q is the heat transfer rate in watts, m-dot is the mass flow rate in kilograms per second, c is the specific heat capacity of the fluid in joules per kilogram kelvin, and ΔT is the temperature difference between the flow and return in kelvin. For water in the relevant temperature range, c is approximately 4,180 J/kg·K and the density is approximately 990 to 1,000 kg/m³, so that one kilogram of water is, for practical purposes, one litre.

Rearranging for mass flow, m-dot = Q / (c × ΔT). The relationship is an inverse proportion: for a fixed thermal output, halving the temperature differential doubles the required mass flow. Taking a single room emitter with a design output of 2 kW as the reference case, a conventional boiler-era differential of 20 K requires m-dot = 2,000 / (4,180 × 20) = 0.0239 kg/s, or approximately 1.43 litres per minute. Reducing the differential to 10 K doubles the requirement to 0.0478 kg/s, or approximately 2.87 litres per minute. Reducing it to 5 K quadruples the original requirement to 0.0957 kg/s, or approximately 5.74 litres per minute. The factor of four is not a design preference but a thermodynamic identity: 20 divided by 5 is 4, and no choice of pipe, pump or controller can deliver the same heat to the room with a lower mass flow without a correspondingly larger temperature drop across the emitter.

It is worth making explicit why the narrow differential is pursued at all. A heat pump's seasonal efficiency is dominated by its return water temperature, since the condenser must reject heat into water that is as cool as possible, and by its flow temperature, which sets the compression ratio. A narrow differential allows the mean water temperature of the emitters to be maintained at a given level with the lowest possible flow temperature, and the lowest return temperature is subsequently obtained from emitters that are sized to release the heat at those temperatures. Compressor manufacturers also typically specify design differentials of 5 K for a number of reasons that include the flow-sensing, defrost and refrigerant-side stability of the unit. The high flow rate is therefore the price paid for efficiency, and for the existing pipework, it is a bill that must be settled in fluid velocity and pressure.

Critical Fluid Velocity and the Erosion-Corrosion Boundary

The velocity of water in a pipe is the volumetric flow rate divided by the internal cross-sectional area. Because the nominal 8 mm and 10 mm descriptions refer to the outside diameter of the tube, the internal diameters must be determined from the wall thickness. For standard microbore copper with a wall of approximately 0.6 mm, an 8 mm tube has an internal diameter of approximately 6.8 mm and a flow area of approximately 36 mm². A 10 mm tube has an internal diameter of approximately 8.8 mm and a flow area of approximately 61 mm². By comparison, a 15 mm tube has an internal diameter of approximately 13.6 mm and an area of 145 mm², and a 22 mm tube has an internal diameter of approximately 20.2 mm and an area of 320 mm².

Applying the 2 kW reference flows, a boiler-era 20 K differential in an 8 mm tube gives a velocity of 0.0239 litres per second divided by 36 square millimetres, which is approximately 0.66 m/s. This is comfortably below the commonly cited ceiling and explains why microbore performed adequately for decades. A 10 K differential raises it to approximately 1.32 m/s, and a 5 K differential raises it to approximately 2.64 m/s. In a 10 mm tube the corresponding figures are 0.39, 0.79 and 1.57 m/s. The microbore branch that was entirely satisfactory at a 20 K differential is therefore carrying fluid at two and a half times the commonly accepted velocity limit when the same room is served by a 5 K heat pump circuit.

The Physical Basis of the 1.0 m/s Guideline

Copper's corrosion resistance in heating water is not an intrinsic property of the metal but the product of a thin, adherent film of cuprous oxide, with a layer of more complex copper corrosion products above it, that forms on the interior surface and isolates the metal from the water. This film is stable only so long as the conditions at the wall are quiescent enough for it to remain attached. The shear stress that the flowing water exerts on the wall is, for turbulent flow, proportional to the density multiplied by the square of the velocity and by a friction factor, so doubling the velocity roughly quadruples the wall shear stress, and the local intensity at bends, tees, reducers and burrs where the flow separates and impinges is many times higher still.

When the local shear stress and turbulence exceed the film's mechanical strength, the film is stripped away. The exposed bare copper re-oxidises rapidly, consuming copper from the wall in the process and forming a new film which is itself removed. This cyclic removal and regrowth is erosion-corrosion, and it differs from conventional uniform corrosion because the rate of metal loss is determined by the hydrodynamics rather than by the chemistry. The characteristic result is localised wall thinning, with horseshoe-shaped or teardrop pits that are aligned with the direction of flow, concentrated immediately downstream of fittings, and which eventually perforate the wall to leave pinhole leaks. Microbore tubes with a wall thickness of only around 0.6 mm have a very small corrosion allowance, so the time from the onset of attack to perforation is correspondingly short.

The effect is aggravated by factors that are common in older systems. Entrained air bubbles collapse at the wall and add mechanical damage. Particulate debris, notably magnetite sludge from steel radiators and scale from old pipework, acts as an abrasive in the flow. Dissolved oxygen and aggressive water chemistry, particularly in systems that have been drained and refilled frequently or that lack an appropriate inhibitor, reduce the protective quality of the film. A system flowing at 2 m/s or more through numerous tight microbore bends, with a history of poor water quality, is thus exposed to the full combination of those effects.

There is also an acoustic dimension. Turbulent flow at high velocity through small passages generates broadband noise, and flow separation at sharp bends and valve seats can excite whistling or hissing that is transmitted along the pipe and into the building structure. In a quiet low-temperature heating system, where the occupants' expectation is of unobtrusive operation, this is among the most frequently reported complaints and a sensitive early indicator that velocities are excessive. It should be noted that design velocity guidance varies with pipe diameter, temperature and the quality of the installation, and that figures between 0.5 and 1.0 m/s are commonly quoted for small-bore copper; the 1.0 m/s ceiling used in this article should be treated as an upper bound for a clean, well-installed system rather than as a target.

The matrix illustrates the central finding. The same 2 kW of delivered heat that produced a comfortable 0.66 m/s in 8 mm microbore at boiler differentials produces a velocity four times higher at a heat pump differential, and a 15 mm tube is required to restore it to the same value. The velocity scales with the inverse of the temperature differential and with the inverse square of the internal diameter, which means that the combined effect of a reduction in differential and a small internal diameter is multiplicative. Figures are indicative, calculated for a single 2 kW emitter leg, and the velocities in shared sections of a manifold serving several radiators will be proportionately higher.

Hydraulic Resistance and Pump Head Constraints

The pressure drop in a straight pipe is described by the Darcy-Weisbach equation: Δp = f × (L / D) × (ρ × v² / 2), where f is the dimensionless Darcy friction factor, L is the pipe length, D is the internal diameter, ρ is the density and v is the mean velocity. Three features of this equation govern the behaviour of microbore. First, the pressure drop is proportional to the square of the velocity. Second, it is inversely proportional to the diameter. Third, because the velocity for a fixed volumetric flow is itself inversely proportional to the square of the diameter, substituting v = 4Q / (π × D²) shows that for a fixed flow the pressure drop varies with the inverse fifth power of the diameter, with a modest adjustment because the friction factor itself depends on the Reynolds number. This is a steep power-law dependence rather than an exponential one, but its practical effect is similarly dramatic: halving the internal diameter increases the pressure drop by a factor of approximately 30.

The friction factor also varies with flow. In the smooth-pipe turbulent regime that applies to copper at these velocities, the Blasius approximation, f = 0.316 × Re^-0.25, gives a pressure drop that scales approximately with the flow rate to the power of 1.75. A quadrupling of flow, which is the penalty associated with a move from a 20 K to a 5 K differential, therefore raises the pressure drop by a factor of roughly 11 rather than 16. For the 2 kW reference leg in 8 mm microbore, the Reynolds number at 5 K is approximately 25,000 for water at around 40 degrees Celsius, giving a friction factor of about 0.025 and a pressure drop of approximately 12 to 13 kPa per metre of straight tube. At the boiler-era differential the corresponding figure is approximately 1.1 kPa per metre. A radiator leg of 15 metres, which is by no means unusual, would therefore present a resistance of roughly 190 kPa, or about 19 metres of water head, before the contributions of bends, tees, valves and the radiator itself are added.

A standard domestic variable-speed circulator typically delivers a maximum head of 5 to 8 metres at low flow, and its head falls as flow increases along the pump curve. The intersection of the pump curve and the system resistance curve defines the operating point. When the system resistance is as high as that of a long microbore leg at 5 K flow, the intersection occurs at a much lower flow than the design requirement. The flow that actually passes is limited by the hydraulic authority of the pump, which is to say that the heat pump cannot circulate its design flow through the microbore at any pump speed available from an integral circulator.

The consequences are practical. The heat pump's flow sensor detects a flow below its minimum threshold, and the unit locks out with a low-flow fault or operates at a reduced differential, with a corresponding increase in the leaving water temperature that undermines efficiency. The compressor's high-pressure protection may trip as the condenser is starved of cooling water. The pump itself, forced to operate at the far left of its curve against a high resistance, runs hot and noisily, and, where the system pressure is low or the water contains dissolved gas, the reduced static pressure at the pump inlet can allow vapour bubbles to form and collapse, which is the mechanism of cavitation and one which damages impellers and bearings. Careful attention to the system's static pressure, to air removal, and to the pump's net positive suction head requirement is therefore an integral part of any microbore design.

The energy cost of overcoming the resistance should also be quantified. The hydraulic power required is the product of the pressure drop and the volumetric flow. For the 190 kPa leg at 0.0957 litres per second, this is approximately 18 watts of hydraulic power, which at a typical small-circulator overall efficiency of 25 to 30% corresponds to an electrical input of 60 to 70 watts to serve a single 2 kW leg. This is approximately 10% of the roughly 670 watts of compressor input needed to produce 2 kW at a coefficient of performance of 3. The pumping penalty for forcing the flow through microbore is thus not negligible in the context of seasonal efficiency, even before the reliability issues are considered.

A further consideration is system water volume. An 8 mm microbore pipe with a 6.8 mm bore holds approximately 36 millilitres per metre, and a 10 mm pipe holds approximately 61 millilitres per metre. Systems that comprise entirely microbore and small-volume radiators have a low total water content, which is a disadvantage for heat pump operation: the compressor needs a sufficient thermal mass in the circuit to avoid short-cycling at low load and to supply the energy for a reverse-cycle defrost without a significant drop in flow temperature. A buffer vessel or additional volume is therefore commonly needed irrespective of the velocity issue.

The Engineering Architecture of Hydraulic Separation

The most robust solution to the velocity and pressure problem is not to force the heat pump flow through the microbore but to decouple the two. Hydraulic separation divides the system into a primary circuit, which serves the heat pump and is designed for its high, constant or modulating flow at a narrow differential, and a secondary circuit, which serves the emitters and is designed for the flow rate that the microbore can tolerate. Each circuit has its own circulator, and the interface between them is arranged so that the two pumps do not influence one another.

Low Loss Headers

A low loss header, also called a hydraulic separator or low loss manifold, is a vertical vessel with primary connections at one end and secondary connections at the other, designed so that the water within it moves at very low velocity. Its function is to create a zone of near-zero pressure differential between the primary flow and return, so that the primary and secondary pumps operate independently. The design rules are well established. The header should be sized so that the internal velocity does not exceed approximately 0.1 to 0.2 m/s at the highest of the primary and secondary flow rates, which in practice means that the vessel's cross-sectional area should be at least three to four times that of the largest connecting pipe, commonly expressed as a header diameter of two to three times the connection diameter. The primary and secondary connections should be arranged with the flows entering and leaving in the same plane, the flow connections at the top and the return connections at the bottom, to permit stratification and to allow the vessel to function as an air separator and dirt trap where provided with the appropriate fittings.

Neutral Points and Close-Coupled Tees

The same principle can be achieved without a vessel by using two tees in the pipework, close-coupled so that the distance between the flow and return connections of the primary and secondary circuits is no more than approximately three to four pipe diameters. The common section between the tees is the neutral point, in which the pressure difference is negligible and so there is no motive force to drive flow from one circuit into the other other than the difference between the two flow rates. Where the primary flow exceeds the secondary flow, the excess passes through the common section from flow to return; where the secondary flow is the larger, the shortfall is made up by recirculation from the return to the flow. A close-coupled tee arrangement is compact and inexpensive, though it lacks the air and dirt separation functions of a purpose-built header, and the geometry must be respected: if the tees are widely separated, a pressure differential develops across them, and the pumps interact.

Thermal Injection and Buffer Configurations

A buffer vessel can be arranged in series, in which case all of the primary flow passes through it, or in parallel, in which case it is connected as a separation device and a thermal store. Parallel or primary-side buffers function as hydraulic separators and also provide the water volume that the heat pump requires for defrost and to prevent short cycling. In an injection arrangement, a mixing or injection valve adds hot primary water to the secondary circuit in the quantity required to maintain the secondary flow temperature, allowing the secondary circuit to operate at a different differential and flow temperature from the primary circuit.

The Trade-Off Inherent in Separation

Hydraulic separation resolves the velocity problem, but it does so at a thermodynamic cost that must be understood. If the secondary flow is lower than the primary flow, which is the design intention when the microbore leg is to be run at a reduced flow, the surplus primary water bypasses the emitters through the header and mixes with the secondary return. The heat pump return temperature is then higher than the emitter return temperature, and the effective primary differential is narrower than the secondary differential. With a flow temperature of 45 degrees Celsius, a secondary differential of 10 K gives an emitter return of 35 degrees Celsius, and the mixture of this return with bypassed water at 45 degrees Celsius produces a heat pump return temperature of approximately 40 degrees Celsius at a 5 K primary differential. The mean water temperature at the emitters is 40 degrees Celsius instead of the 42.5 degrees Celsius that the equivalent 5 K secondary circuit would deliver, so the emitters output proportionally less heat, and compensation by a higher flow temperature reduces the coefficient of performance. Conversely, if the secondary flow exceeds the primary, the header recirculates hot water from the flow to the return of the secondary, reducing the secondary supply temperature below that of the heat pump leaving temperature, with a corresponding loss of emitter output.

The key design rule is therefore that the primary flow must be at least equal to the secondary flow, so that the supply temperature is preserved at the emitters, and that the bypassed flow, together with the associated penalty, is minimised by matching the circuits as closely as the microbore limits allow. In practice this means reducing the number of radiators served per microbore manifold branch, keeping leg lengths short, selecting emitters with a higher output per unit of flow, and where possible converting the principal feeds to 15 mm or 22 mm and retaining microbore only for the final short connection to each emitter. Because microbore velocity is highest in the shared sections that serve multiple radiators, replacing those sections is generally the most efficient step in a partial conversion.

Specialised Circulation Arrays

The secondary circuit's pump must be selected for the head requirement of the microbore, which may exceed the capability of a standard domestic circulator. High-head variable-speed circulators, with a maximum head of 10 to 12 metres, are available, and may be configured with a proportional-pressure or constant-pressure control curve. A twin-head circulator comprises two pump heads in a single housing, which can operate in duty-standby mode for resilience, or in parallel for doubled flow at a given head. Pumps connected in series increase the head at a given flow. These configurations follow the pump affinity laws: the flow is proportional to the shaft speed, the head is proportional to its square, and the power is proportional to its cube, so that a modest increase in speed to overcome a resistance carries a steep penalty in electrical consumption. It is for this reason that increasing the pump capacity should be regarded as a mitigation and not a cure: it enables an existing microbore system to function, but it does so by spending electrical energy to overcome a resistance that a larger bore would not present at all.

Diagnostic Procedure and Specification Guidance

A systematic assessment of an existing microbore installation should begin by establishing the pipe material, the internal diameters, the circuit topology and the length of each leg. For each emitter the design heat output at the design flow temperature should be calculated and converted to a flow rate for the intended differential, from which the velocity and pressure drop follow. Where the velocity exceeds approximately 1.0 m/s in any section, the options are, in rough order of increasing cost, to widen the emitter differential by accepting a higher flow temperature, to reduce the load per leg by rebalancing or splitting circuits, to replace the shared and highest-velocity sections with 15 mm or 22 mm pipework, or to replace the microbore entirely. The cost of conversion varies widely with the construction of the property and the accessibility of the pipe routes, and it should be established by survey rather than assumed, but it is generally lower than the cost of a failed installation and the remedial work associated with perforated pipework.

Whatever route is chosen, the water quality of the circuit must be addressed. Before commissioning, the system should be cleaned and flushed in accordance with the relevant standards and guidance, a magnetic and particulate filter should be fitted on the return to the heat pump, an inhibitor appropriate to the system's materials should be dosed, and the water chemistry should be verified. Cleaning does not change the velocity limits, but it removes the abrasive and corrosive agents that convert a marginal velocity into a failure. Commissioning should include verification of the actual flow rate with a calibrated flow measurement, confirmation of the differential at design output, and a record of the pump settings and the hydraulic arrangement.

The conclusion for specifiers is that a heat pump can be integrated with a legacy microbore network, but it cannot be integrated unchanged. The velocity and pressure drop arising from a 5 K differential in 8 mm or 10 mm tube are incompatible with the pipework's physical limits and the capability of a standard circulator, and the choice is between a partial replacement of the pipework, a hydraulically separated architecture with the accepted efficiency trade-offs, or a combination of both. Each option can be quantified at the design stage, using the equations in this article, and the decision should be based on that quantification rather than on the assumption that an existing system which served a boiler will continue to serve a heat pump.