# White paper on RenewX heat exchangers

Dariusz Mikielewicz, Jarosław Mikielewicz, Piotr Doerffer

### 1. Introduction

Air source heat pumps use air as a heat source to evaporate the working fluid in evaporator of the thermodynamic cycle. The working fluid removes that heat together with the work supplied to compressor at required temperature by operational parameters. The research activities carried out worldwide aim at development of the effective and at the same time with a smaller footprint evaporator for transfer of heat from supply air to the working fluid, which is easy to manufacture. In such case heat transfer from air to the channel walls is significantly poorer than that from the channel wall to the working fluid. Hence the objective of activities in the project were focused on intensification of heat transfer from air to the evaporator channels.

Air-source heat pumps (ASHP) are typically based on a basic brazed fin and tube heat exchanger (evaporator and condenser). As a result of the relatively low efficiency of the brazed interface, many fins are required which greatly increase the size, weight and cost of the unit. Simultaneously, a very large density of fins requires a large fan to force air through a wide heat exchanger. The fan is therefore noisy, requires maintenance and consumes approximately 5 to 8% of the energy demand of the overall system.

In addition, ASHP are typically installed outside which makes them susceptible to frost formation. Frost will build up on the evaporator from temperatures as high as 5°C. In order to cope with frost formation the main body of other research is to reduce the size of the evaporator (and hence the volume of aluminum) by enhancing the local air-side heat transfer coefficient. This local intensification is achieved through a progressive reduction of the air-side cross-sectional area, which results in an increase of the mean velocity of the air stream. The main drawback of the implementation of heat transfer enhancement techniques is an intrinsic increase of the reversible and irreversible components of the airside pressure drop of the evaporator.

Wall to fluid heat and mass transfer are influenced by the thermal and hydrodynamics boundary layers. In the recent years, numerous studies were conducted to better control and/or to increase parietal heat transfer. The general objective now is to maximize the overall heat transfer coefficient U and therefore minimize heat transfer area A. This will lead to the smallest physical heat transfer device and the lowest cost.

Heat transfer intensification is a philosophy that aims at achieving reduced size of heat transfer equipment and the associated benefits (especially improving heat transfer performance). Thus, the product of intensification techniques is achieving any, or a combination, of the following :

- reduction in size of the heat exchanger for a given duty;
- increase in capacity of an existing heat exchanger;
- reduction in approach temperature difference;
- reduction in pumping power.

Intensified heat transfer will increase efficiency, which will lead to energy conservation and reduced costs. However, some techniques may increase power consumption, either directly or due to increased flow resistance through the exchanger.

### 2. Air side enhancement techniques

Passive heat transfer intensification techniques include mainly:

- enhanced surfaces (fins), see Fig. 1;
- rough surfaces;
- displaced enhancement devices;
- swirl flow devices;
- coiled tubes.

There are many recent approaches in this area as which result in:

- increasing fluid turbulence,

- secondary flow development,

- disruption of the thermal boundary layer,

- and increasing the heat transfer surface area,

- application of minichannels.

Typical fin spacings are 300 to 800 fins/m. Due to their small hydraulic diameter and the low density of gases, these surfaces are usually operated in the Reynolds number range from 500 to 1500. As a result, plate-fin or tube-fin enhancement geometries must be effective in the low Reynolds number regime. For example, surface roughness has been shown to promote heat transfer in the turbulent regime, but it does not provide appreciable enhancement in the lower Reynolds number range.

Fig. 1. Types of extended surfaces - fins

Universal types of heat exchangers incorporating plate–fins or fin–tubes are designed for the applications in which industrial processes and systems need to be resolved, and for these types the flow is channeled between plates. Various surface extensions in the form of fins and turbulators are used to improve air side heat transfer. Amongst the different fin types can be mentioned, see also Fig. 2:

- Louvered fins (modified from Chang et al., 2000),
- Wavy fins (from Wang et al., 1999c),
- Slit fins (modified from Kim and Jacobi, 2000),
- Rectangular offset strip fins (from Manglik and Bergles, 1995).

Fig. 2. Various fin types and geometrical parameters: (a) Louvered fins, showing a cross section of the louvers, a cross section of tube, and an end and front view of a serpentine louvered fin between two flat tubes; (b) Wavy fins, (c) Slit fins, (d) Rectangular offset strip fins

Longitudinal vortex generators (LVGs) can be mounted on channel walls to generate longitudinal vortices which create a secondary flow and disturb the boundary layer growth, thereby enhance the heat transfer between the flowing fluid and channel walls. Examples of LVG are shown in Fig. 3. That solution has been selected for the activities in the project.

Fig. 3. Examples of longitudinal vortex generators

### 3. Evaporators of heat pumps

Usually the thermal-hydraulic calculation of the relevant heat exchangers is accomplished by commercial computational fluid dynamics (CFD) software. In present calculations the ANSYS-Fluent code was used. The standard approach to use the CFD code is to appropriately define the calculation mesh and solve the governing equations for the assumed geometry. The mathematical model, which is accurate for each studied geometry of the micro-channel heat exchanger consist of equations describing the conservation of mass, momentum and energy and appropriate closure equations. Set of equations needs to be supplemented by the model of turbulent kinetic energy evolution and evolution of the dissipation of kinetic energy. In the classical CFD model, which has been used, there haven't been taken into account the equations describing non-equilibrium formation of the new phase and that is why the classical model had to be supplemented by author's own equations

### 3.1. Heat pump cycle

In the present research R134a has been selected as a working fluid for the heat pump cycle. Obtained parameters of the heat pump cycle have been shown in tables 1 and 2. In table 1, the evaporator parameters are presented with assumption, that the minimum temperature difference between air and working fluid was equal to 5K, and air inlet temperature is equal to 7°C. In turn, in table 2, there have been shown the results of the condenser analysis, in which the water is being heated from 16°C to 60°C. There has been also kept fluids temperature difference at the level of 5K. For the heat pump cycle parameters given in tables 1 and 2, there has been calculated the COP value equal to 3.28.

Table 1. Evaporator parameters for water outlet temperature t_{out}=60°C

| t_{in} | t_{out} | x_{in} | x_{out} | p | m | m | V_{in} | V_{out} |

| [°C] | [°C] | [-] | [-] | bar | kg/s | kg/h | m^{3}/min | m^{3}/min |

Air | 7 | 3 | 1 | 1 | 1 | 1,71 | 6150 | 80.633 | 79.473 |

R134a | -2 | 2 | 0.464 | 1 | 2.68 | 0.0623 | 224.28 | 0.133 | 0.287 |

Table 2. Condenser parameters for water outlet temperature t_{out}=60°C

| t_{in} | t_{out} | x_{in} | x_{out} | p | m | m | V_{in} | V_{out} |

| [°C] | [°C] | [-] | [-] | bar | kg/s | kg/h | m^{3}/min | m^{3}/min |

Air | 16 | 60 | 0 | 0 | 4 | 0.05 | 180 | 0.00352 | 0.00363 |

R134a | 78 | 60 | 1 | 0 | 16,78 | 0.0623 | 224.28 | 0.04866 | 0.00381 |

The evaporator design is based on the finned-tube heat exchanger with longitudinal vortex generators. The design specification is as shown in Table 3.

Table 3. Design specification of heat pump evaporator

Parameter | Value |

Duty | 4kW |

Air temperature drop | 5°C |

Heat exchanger minimum approach temperature | 5°C |

Working fluid | R134a |

Working fluid superheat | 0°C |

The heat balance for the air and working fluid yields the following stream flow rates:

- air mass flow rate: 0.8 kg/s,
- refrigerant mass flow rate: 0.02kg/s.

The duty of the heat exchanger can be determined from the formula:

In equation (1) Q̇ denotes the heat exchanger duty (W), U is the overall heat transfer coefficient (W/m^{2}K), Aref is the area of the refrigerant side (m^{2}), F_{T} is the mean temperature correction factor (1, for phase change heat exchangers) and ΔT_{LM} is the log mean temperature difference between the two feeds (7.21K for a minimum approach temperature of 5, air temperature drop of 5 and an isothermal working fluid).

The overall heat transfer coefficient is found using:

In equation (2) h_{ref} is the refrigerant side heat transfer coefficient (W/m^{2}K) and h_{air,corrected} is the air

side heat transfer coefficient corrected to account for the extended surface area on the air side

(W/m^{2}K). The corrected air-side heat transfer coefficient is calculated as follows:

In equation (3) A_{ratio} is the ratio between the air-side and refrigerant-side area and η_{fin} is the fin efficiency (assumed to be 95%).

### 3.2. Plate-fin heat exchanger

In case of the plate fin exchanger it was assumed that the working fluid passage will consist of 7 channels, as shown in fig. 4. Such selection has been done with respect to the issue of appropriate fluid distribution into respective channels. Calculations have been obtained by means of CFD calculations applied to a repeatable segment of the plate-fin arrangement. The computational mesh has been established using the Gambit software. Ansys CFX has been used as a modelling tool. Calculated value of the air heat transfer coefficient is 53W/m^{2}K. In Fig. 4 there have been presented fundamental dimensions of the microchannel heat exchanger segment. Its analysis has been the basis for specyfying final size of the whole heat exchanger. Summary of plate-fin HX design (without intensification) for 5x2.5mm channels with R134a as working fluid obtained by Ansys CFX calculations are presented in Table 4.

Fig. 4. Channel segment dimensions for the plate-fin heat exchanger

Fig. 4. Channel segment dimensions for the plate-fin heat exchanger

Table 4. Summary of plate-fin HX design (no intensification) for 5x2.5mm channels with R134a as working fluid obtained by Ansys CFX calculations

**Type** | Plate-fin, 5mmx2.5mm |

**Flow Configuration** | Cross-flow |

**Fin Type** | Plain, rectangular |

**Fin Pitch** | 2 mm |

**Fin Length** | 23,3mm |

**No. Fin Banks** | 16 |

**Plate Type** | Flat, divided into sub-channels |

**No. Plate Channels** | 15 |

**No. Sub-Channels per Plate** | 7 |

| |

**Air-Side Area** | 6.114 m^{2} |

**R134a-Side Area** | 0.6426 m^{2} |

**Area of Fins** | 5.675 m^{2} |

**Required Fin Efficiency** | 89 % |

**Over-Size (Air-Side)** | 0 % |

**Over-Size (R134a-Side)** | 0 % |

**Pressure Drop (Air-Side)** | 0.039 kPa |

**Pressure Drop (R134a-Side)** | 0.260 kPa |

| |

**Fan Size Required** | 0.099 kW |

**Module duty** | 1.790kW |

#### 3.2. Tube fin heat exchanger

In case of tube-fin heat exchanger the geometry studied in presented in Fig. 5. The heat exchanger design model is used to size the evaporator. To obtain the air-side heat transfer coefficient, the following data (Figure 6) can be used.

Fig. 5. Schematic of the tube-fin heat exchanger and a channel segment dimensions

Own calculations using ANSYS Fluent software, related to evaluation of various turbulators and pitches has been collected in Fig. 6. Hence, the data for the air side heat transfer coefficient (uncorrected, h_{air}) can be extracted using the following method:

- Study model geometry to obtain refrigerant channel and air ratio data,
- Use relevant correlation to estimate two-phase flow heat transfer coefficient h
_{ref}, - Use overall heat transfer coefficient equation to estimate h
_{air}.

The air-side heat transfer coefficient as a function of Reynolds number for a 4mm fin pitch and 24mm width air ducts is therefore obtained and displayed in Fig. 7. The air-side heat transfer coefficient is therefore taken as 132 W/m^{2}K for a Reynolds number of 2300 (which is adhered to in the final design). In order to calculate the area ratio, the heat exchanger geometry was set to 4mm fin pitch, 24mm air duct width and 5mm tube pitch (longitudinal to air flow).

Fig. 8a. Section of finned-tube structure (not showing LVG enhancement)

Fig. 8. b) 4mm section of one tube and associated fin area (side view; not showing LVG enhancement)

Fig. 8. c) section of one tube and associated fin area (top view; not showing LVG enhancement)

The tube-fin heat exchanger geometry (Table 5) satisfies the required area and practical pressure drop constraints.

The following thermal-hydraulic uncertainty exists in the design presented in Table 5:

- The air-side pressure drop does not account for the LVG enhancement. This was impossible to quantify without friction factor vs Reynolds number data. However, as the pressure drop for plain fins is so low (0.006% of atmospheric pressure) it is not expected that the real pressure drop would be detrimentally high. Initial testing of the prototypes suggests that this assumption is valid.
- The air flow length is 0.06m. In the original studies from WP1, the air flow length is 0.04m. It is uncertain as to whether the flow disruption caused by the LVG (and subsequent heat transfer enhancement) will be effective at this length. Therefore, it is worth studying (experimentally or by modelling) the use of two LVGs with one positioned at the flow entrance and one positioned at the centre point of the flow passage.
- The overall heat transfer coefficient did not consider fouling of the air side which may be significant. Sinott (2006) give a typical air fouling factor of 0.00025m2K/W. This would reduce the value of the overall heat transfer coefficient to 925 W/m2K a loss of 23% which would roughly correspond to an increase in required area of 23%. However, the fouling factor is dependent on a number of parameters including local air velocity, turbulence and air cleanliness. It is noted that the value given by Sinott (2006) is for typical process industry applications where the air may particulate-heavy, and the heat exchangers are of convention laminar (on the air-side) flow design. Hence, this fouling factor may not be valid for this particular design as the air condition may be different and fouling may be negated due to the increased air velocity and turbulence generated by the LVG enhancement of the fins.

**Table 5. Suggested tube-fin evaporator design summary**

**Refrigerant side area** | 0.483 m^{2} |

**Fin area** | 5.12 m^{2} |

**Total air side area** | 5.73 m^{2} |

**Area Ratio** | 11.85 |

| |

**Tube outer diameter** | 0.005 m |

**Tube thickness** | 0.0005 m |

**Tube inner diameter** | 0.004 m |

**Tube length per pass** | 0.428 m |

**No. tubes per pass** | 6 in-line |

**No. passes** | 5 |

**No. Sections** | 3 |

**Tube material** | Copper |

| |

**Fin width** | 0.428 m |

**Fin depth** | 0.06 m |

**Fin features** | Continuous with LVG between "bundles" of in-line tubes |

**No. fins** | 107 |

**Fin material** | Aluminium |

| |

**Air-side Reynolds number** | 2200 |

**Average refrigerant-side Reynolds number** | 16600 |

| |

**Air side pressure drop** | 5.91 Pa ( <1% of atmospheric pressure) |

**Refrigerant side pressure drop** | 1590 Pa ( <1% of inlet pressure at 0oC; excluding U-bends and headers) |

**Heat exchanger duty** | 4kW |

Material selection for tubes went towards aluminum tubes due to the following lessons learnt concerning manufacturing and materials:

- If a heat exchanger design was built entirely from aluminium (tubes and fins) then various manufacturing issues were noted by Powerfin ltd (manufacturer of the prototypes) which are summarized as follows:
- Difficulty in creating/buying U-bends constructed from aluminium.
- Difficulty in sealing aluminium U-bends to the tubes.

- Therefore, it is proposed that copper is used for the tube material in this case. Aluminium is still used for the fins.
- This has a further advantage in manufacturing HEX-1 from the air-solar system. Aluminium tubes would not be compatible with water due to corrosion concerns. Hence, copper must be used in HEX-1. If both units are constructed from copper tubing, it is anticipated that manufacture of the two heat exchangers will be cheaper and easier.
- It is not anticipated that there will be any issues in joining the aluminum fins to copper tubes.

There are two items of interest in terms of the manufacture of this heat exchanger. Firstly, incorporating the linear vortex generators needs to be carried out in a cost effective manner, to enable this additional manufacturing step to be considered commercially viable. The second manufacturing challenge is to identify a method of attaching the fins to the tubes which allows for higher heat transfer compared to mechanical attachment. Traditionally the fins would be mechanically fitted by deforming the tubes. However there remains a thermal contact resistance if only a mechanical attachment method is used.

#### 3.3. Microchannel heat exchanger

The core part of the MCHE consists of the set of stainless steel shims with 67 electrochemically etched channels of the length of 17mm, and width of 300μm and depth equal to 100μm. The number of such plates depends on the duty of the proposed HX. The HX channels have been distributed with the spacing of 1mm. For the supply and removal of fluids there have also been etched fluid supplying and outlet collectors of the depth equal to 200μm and the uniform width of 4mm. Outline of the described single shim of the proposed HX has been presented in Fig. 8.

Fig. 8. Dimensions of the stainless steel shim

Fig. 9. Photograph of a shim

View of the single stainless steel shim, featuring the channel dimensions of 0.3x0.1mm, through which there have been realised micro-flows, has been given in Fig. 9. According to the plates layout, the channels have been supplied with a cold or hot medium, as it has been clearly shown in Fig. 10.

Fig. 10. Expanded view of the assembly of shims

The modular construction allows to increase the capacity of the HX, in a quite simple way, through addition of the heat transfer area in the form of a shim. In the system, which has been explored, every cycle consists of four shims (each shim has 67 channels). The stainless steel shims will be assembled into the heat exchanger, including the supplying/outlet collector.

In calculations of the dimensions of the HX the following details of the plate size have been assumed:

- Channel length L=17 mm
- Channel cross-section a=0.3 mm b=0.1 mm
- Channel side wall (fin) thickness d
_{s},_{water}=0.7 mm (1 mm pitch) - Wall thickness between R134a and water d
_{R}=0.2 mm - Number of water channels in one layer nW1=2x67=134 (2 rows in one layer)
- Number of refrigerant channels in one layer n
_{R1}=2x67=134 (2 rows in one layer)

Calculations for the case of condenser have been accomplished for the initial refrigerant temperature (superheated conditions) T_{R},in= 84.72°C at corresponding saturation temperature T_{R},_{x=1}=65°C, p_{R,in}=18.91371 bar. In result the dimensions for stainless-steel condenser have been obtained such as: width W_{x}=6.70 cm, height W_{y}=11.94 cm, length 2xL=2x1.7 cm=3.4 cm. Specific results for the case of the microchannel HX along with the heat transfer coefficients in the condensations zone are presented in Table 6.

**Table 6. Specific results for the case of the microchannel HX, channel spacing 1mm**

L mm | n_{W} - | n_{R} - | n_{DL} - | W_{x} cm | W_{y} cm | V dm^{3} | m kg | h_{W} W/m^{2}K |

17 | 26666 | 26666 | 199 | 6.70 | 11.94 | 0.27 | 1.96 | 15209 |

h_{R} W/m^{2}K | U W/m^{2}K | Re_{W} - | G_{R} kg/m^{2}s | Δp_{W} Pa | Δp_{R} Pa |

5192 | 3799 | 56 | 79.3 | 2197 | 163 |

L – water or refrigerant or water channel length

nW – total number of water channels

nR – total number of refrigerant channels

nDL – number of double layers (of refrigerant and water channels)

Wx – width of condenser’s core

Wy – height of condenser’s core

V – volume of condenser’s core

m – mass of condenser’s core

hW – heat transfer coefficient at water side

hR – heat transfer coefficient at refrigerant side

U – overall heat transfer coefficient in condensation zone

ReW – water flow Reynolds number

GR – refrigerant flow mass velocity

ΔpR – refrigerant pressure drop

ΔpW – water pressure drop

The suggested design is presented in Table 7.

**Table 7. Suggested microchannel plate heat exchanger design summary**

**Type** | Microchannel |

**Flow Configuration** | Counter-flow |

**Channel Type** | Plain, rectangular |

**Channel Pitch** | 1.0 mm |

**Channel Length** | 17 mm |

**No. Channel Banks** | 2x398 (water + refrigerant) |

**No. Channels per Bank** | 67 |

**Shim type** | Flat, with microchannels |

**Shimthickness** | 0.3 mm |

**No. Channel shims** | 2x199 (water + refrigerant) |

**No. Banks per shim** | 2 |

| |

**Water-side area** | 0.362 m^{2} |

**R134a-side area** | 0.362 m^{2} |

**Over-size (Water-Side)** | 0 % |

**Over-size (R134a-Side)** | 0 % |

**Pressure drop (Water-side)** | 2.197 kPa |

**Pressure drop (R134a-side)** | 0.163 kPa |

| |

**Module duty** | 10 kW |

### 4. Conclusions

Two original designs of heat exchangers have been manufactured and tested, namely the tube-fin heat exchanger to operate as evaporator (Fig. 11), and a microchannel heat exchanger consisting of diffusion bonded stainless steel shims to operate as condenser.

Systematic tests confirmed their consistency with the thermal-hydraulic models.

Lessons learned from the project lead to recommendations to further investigation and optimisation of air-to-liquid heat exchange by means of:

- Accelerated air flow (progressive reduction of the air-side cross-section area)
- Non-symmetrical refrigerant channels
- Maldistribution in supply channels

Fig. 11. Comparison of the size of typical commercial heat pump evaporator and the one devised within RenewX project

### References

- Chang, Y. J., Hsu, K. c., Lin, Y. T. and Wang, C. C. A Generalized Friction Correlation for Louver Fin Geometry," Int. J. Heat Mass Transfer, Vol. 43, pp. 2237-2243, 2000.
- Jacobi, A. M. and R. K. Shah, Air-Side Flow and Heat Transfer in Compact Heat Exchangers: A

Discussion of Enhancement Mechanisms, Heat Transfer Engineering, 19:4, 29-41, 1998. - Joardar, A., Jacobi A. M., Impact of Leading Edge Delta-Wing Vortex Generators on the Thermal

Performance of a Flat Tube, Louvered-Fin Compact Heat Exchanger," International Journal of

Heat and Mass Transfer, 48, 1480-1493, 2005. - Hwang S.W., Kim D.H., Min J.K., Jeong J.H. CFD analysis of fin tube heat exchanger with a pair of

delta winglet vortex generators, Journal of Mechanical Science and Technology, Vol. 26 (9), 2949-2958, 2012. - Kew P.A., Reay D.A., Compact/micro-heat exchangers - Their role in heat pumping equipment, Applied Thermal Engineering 31, 2011.
- Kim, G. J and Jacobi, A. M, Condensate Accumulation effects on the Air-Side Thermal Performance of Slit-Fin Surfaces," CR-26, ACRC, University of Illinois at Urbana-Champaign, IL, 2000.
- Manglik R.M., Bergles A.E., Heat transfer and pressure drop correlations for the rectangular offset-strip-fin compact heat exchanger, Exp. Thermal Fluid Sci., 10, 171-180, 1995.
- Sinott R.K., Coulson and Richardson's Chemical Engineering Series, Volume 6 - Chemical Engineering Design, 4th Edition, Knovel, 2006.
- Wang, C. C., Chi, K. Y., Chang, Y. J. and Chang, Y. P. Experimental Study of Heat Transfer and Friction Characteristics of Typical Louver Fin-and-Tube Heat Exchangers, Int. J. Heat Mass Transfer, Vol. 41(4-5), pp. 817-822, 1998.
- Yan, W. M., Sheen, P. J., Heat Transfer and Friction Characteristics of Fin-and Tube Heat

Exchangers, Int. J. Heat Mass Transfer, vol. 43, p. 1651-1659, 2000.