**2. Materials and Methods**

The selected air handling unit (AHU) is located on the flat roof of a shopping center, located in the city of Eger in Hungary, which has supplied fresh air to the back-o ffice and storage area of a shop since 2017.

#### *2.1. Description of the Investigated Central Ventilation System*

The main air handling components of the system are an air-to-air rotary heat wheel, a mixing box element, and direct expansion cooling/heating (DX) coil connected to a variable refrigerant volume outdoor unit. Figure 1 shows the elements of the investigated AHU.

**Figure 1.** Photo from the investigated air handling units (AHU).

The specification of the AHU can be seen in Table 1.

**Table 1.** Specification of the investigated AHU [19].


Figure 2 shows the outdoor unit which is connected with refrigerant pipes to the DX coil and is located in the AHU.

**Figure 2.** Photo from the outdoor unit.

Technical data of the unit can be seen in Table 2.

**Table 2.** Technical data of the outdoor unit [19].


Table 3 shows the specification of the air-to-air recovery heat wheel in the cooling period.

**Table 3.** Specification of the investigated heat wheel [19].


#### *2.2. Description of the Developed Measurement System*

In total, six temperature and relative humidity sensors, three CO2 sensors and three air velocity sensors were placed in the inlet and outlet section of each air handling element and an electricity energy meter was installed in the outdoor unit. The placement of the measurement points can be seen in Figure 3. The technical data of the installed sensors and instrument can be read in Table 4.

**Figure 3.** Schematic diagram (**a**) for the placement of the measurement points on the real operating AHU and (**b**) the AHU without a heat recovery operation assumption.


**Table 4.** Specification of the sensors and instrument.

The recording of the measured data took place in an hourly period. With regards to the measurement accuracy, temperature sensors are normally used with a ±0.4 ◦C accuracy, humidity sensors with a ±3% accuracy, an air velocity sensor with a ±0.2 m/s accuracy, carbon dioxide sensors with a ±50 ppm accuracy, and an electric energy meter with a 1% of full scale accuracy. Among the monitoring air handling data, the air temperature and relative humidity data of the inlet and outlet sections of the DX cooling coil, energy recovery unit, and outdoor were used to investigate the energy performance and thermal behaviour of these air handling elements in the AHU in the cooling season.

The specification of the sensors and electricity energy meter used for monitoring of the investigated AHU can be seen in Table 4.

For the monitoring and recording of the various air condition parameters and the electrical energy consumption of the outdoor unit, the CentraLine Building Management System (BMS) software (version 2019) solution from Honeywell was implemented on a central server. Figure 4 shows a picture of the target building, along with a representative BMS screen for the investigated air handling unit.

**Figure 4.** A screenshot of the investigated AHU in the Building Management System (BMS).

Access to the BMS software was remotely enabled. Within this technical context, the necessary data were collected for this field study. Data were collected online at hourly intervals, saved, and stored on a computer from a distance.

#### **3. Evaluation of the Data Recorded**

To investigate the energy performance of the AHU, using the measurements, the following mathematical approaches were implemented.

#### *3.1. Calculation Formulas for Measured Data Evaluation*

Using the measured air temperature and relative humidity data, the specific humidity could be calculated to obtain the enthalpy of the air. To achieve this, the water vapor saturation pressure ( *Pws*) was first calculated with Equation (1) [20]:

$$P\_{\text{ws}} = A \cdot 10^{\left(\frac{\text{mt}}{t + t\_n}\right)} \cdot 100,\tag{1}$$

where *Pws* is the saturation pressure of the water vapor in Pa; t is the air temperature in ◦C; and A, m, and *tn* are constant values in -. Since the temperature range during the measurements was between −20 and +50 ◦C, the constant values (in 0.083% maximum error) were as follows: *A* = 6.116441; *m* = 7.591386; *tn* = 240.7263 [20]. The constant value of 100 in Equation (1) represents the conversion of the saturation pressure of water vapor from hPa to Pa.

To obtain the moisture content, the partial pressure of water vapor in the air at a given relative humidity was also calculated with Equation (2) [20]:

$$P\_w = P\_{ws} \cdot \frac{RH}{100} \,\tag{2}$$

where *Pw* is the partial pressure of water vapor in Pa and RH is the relative humidity of the air in %. The constant value of 100 in Equation (2) represents the conversion of relative humidity from % to -.

The moisture content was calculated with Equation (3) [21]:

$$\text{tax} = 0.622 \cdot \frac{P\_w}{P\_o - P\_{w\prime}} \tag{3}$$

where *x* is the moisture content of the air in kg/kg, *Po* is the barometric pressure in Pa, and the 0.622 constant value is the molecular weight ratio of water vapor to dry air.

The enthalpy was calculated with Equation (4) [21]:

$$h = c\_{pu} \cdot t + \mathbf{x} \cdot (c\_{pw} \cdot t + 2500),\tag{4}$$

where *h* is the enthalpy of the air in kJ/kg, *cpa* is the specific heat of air at constant pressure in kJ/(kg· ◦C), *cpw* is the specific heat of water vapor at constant pressure in kJ/(kg· ◦C), and the constant value of 2500 represents the evaporation heat in kJ/(kg· ◦C).

## *3.2. Formulas for Energy Calculations*

Considering the fact that there is balanced ventilation, the e ffectiveness values of the heat wheel were determined from the air temperature measured values using Equation (5) [22,23]:

$$\varepsilon\_s = \frac{(t\_{HWS} - t\_o)}{(t\_{HWE} - t\_o)'} \tag{5}$$

where ε*S* is the real sensible e ffectiveness of the heat wheel given by the measured data in -, *tHWS* is the air temperature in the supply outlet section of the heat wheel in ◦C, *tHWE* is the air temperature in the exhaust inlet section of the heat wheel in ◦C, and *to* is the ambient air temperature which is equal to the air temperature in the supply inlet section of the heat wheel in ◦C.

To ge<sup>t</sup> information about the seasonal energy performance of the heat recovery during the cooling period, the average of the sensible e ffectiveness was calculated with Equation (6):

$$
\overline{\varepsilon}\_{s\\_AV} = \frac{\sum\_{i=1}^{n} \varepsilon\_{s\\_i}}{n},
\tag{6}
$$

where ε*s* \_ *AV* is the average of the sensible effectiveness of the heat wheel given by the measured data in the cooling season in - and n is the number of measurements.

The maximum value of the sensible effectiveness was also analyzed during the whole cooling season, which was calculated with Equation (7):

$$
\varepsilon\_{\text{s\\_MAX}} = \text{MAX}\left(\varepsilon\_{\text{s\\_j}}, \dots, \varepsilon\_{\text{s\\_ll}}\right),
\tag{7}
$$

where ε*s* \_ *MAX* is the maximum value of the sensible effectiveness of the heat wheel given by the measured data in the cooling season in -.

To calculate the energy saving of the heat wheel in the cooling season, Equation (8) was used:

$$Q\_{\rm HWS\\_sorted} = \dot{m}\_{\rm s} \cdot (h\_o - h\_{\rm HWS}) \cdot \tau\_{\prime} \tag{8}$$

where *QHW\_saved* is the energy saving of the heat wheel in kWh; .*ms* is the air mass flow rate delivered by the fans in kg/s, which is calculated by the multiplication of the measured air velocity in m/s and the internal cross-section of air duct in 0.7398 m<sup>2</sup> and approached a 1.2 kg/m<sup>3</sup> constant air density; ho is the ambient air enthalpy which is equal to the air enthalpy in the supply inlet section of the heat wheel in kJ/kg; *hHWS* is the air enthalpy in the supply outlet section of the heat wheel in kJ/kg; and τ is the time in hours. The average air volume flow rate was evaluated as 1060 m<sup>3</sup>/<sup>h</sup> during the cooling season.

To calculate the cooling energy consumption of the DX coil, Equation (9) was used:

$$Q\_{DX\\_HW} = \dot{m}\_\text{s} \cdot (h\_{HWS} - h\_{DX}) \cdot \tau\_\prime \tag{9}$$

where *QDX\_HW* is the cooling energy consumption of the DX coil in kWh, and hDX is the air enthalpy in the supply outlet section of the DX coil in kJ/kg, which is equal to the supply air condition.

In order to investigate more the energy saving impact of the heat wheel on the DX coil, the cooling energy consumption of DX coil was also determined by Equation (10), neglecting the air-to-air rotary heat wheel operation, when the DX coil directly cools the hot ambient air to the supply air conditions.

$$Q\_{DX\\_NO\\_HW} = \dot{m}\_{\mathbb{S}^\cdot} (h\_{\mathbb{S}} - h\_{DX}) \cdot \tau\_\prime \tag{10}$$

where *QDX\_WO\_HW* is the cooling energy consumption of the DX coil without the heat wheel operation in kWh.

The calculated electric energy consumption of the outdoor unit was calculated with Equation (11):

$$P\_{VRV\\_HW} = \frac{Q\_{DX\\_HW}}{EER} \, , \tag{11}$$

where *PVRV\_HW* is the calculated electric energy consumption of the outdoor unit with the heat wheel operation in kWh, and EER is the energy efficiency ratio, given by the producer in -.

Moreover, the real electric energy consumption of the outdoor unit (*PVRV\_HW\_M*) was also measured during the cooling season, in order to see the agreemen<sup>t</sup> between values of the measured data and calculations using the recorded air condition parameters (*PVRV\_HW*). The difference between the measured and calculated electric energy consumption was determined with Equation (12):

$$
\Delta P\_{VRV\\_HW} = P\_{VRV\\_HW\\_M} - P\_{VRV\\_HW\\_t} \tag{12}
$$

where Δ*PVRV\_HW* is the difference between the measured and calculated electric energy consumption of the outdoor unit with the heat wheel operation in kWh.

\_

\_

\_

The rate of deviation of the measured and calculated electric energy consumption of the outdoor unit related to the measured data was calculated with Equation (13):

$$
\Delta P\_{VRV\\_HW\\_REL} = \frac{\Delta P\_{VRV\\_HW}}{P\_{VRV\\_HW}} \cdot 100\_\prime \tag{13}
$$

where Δ*PVRV\_HW\_REL* is the rate of deviation of the measured and calculated electric energy consumption of the outdoor unit in %.

The electric energy consumption of the outdoor unit without the heat wheel operation was calculated with Equation (14):

$$P\_{VRV\\_WO\\_HW} = \frac{Q\_{DX\\_WO\\_HW}}{EER} \,\, \,\tag{14}$$

where *PVRV\_WO\_HW* is the electrical energy consumption of the outdoor unit without the heat wheel operation in kWh when it directly cools the hot ambient air to the supply air conditions via the DX coil during the cooling season.

The energy saving of the heat wheel in terms of the electric energy consumption of the outdoor unit was calculated with Equation (15):

$$
\Delta P\_{\rm VRV\\_HW\\_saved} = P\_{\rm VRV\\_WO\\_H\mathcal{W}} - P\_{\rm VRV\\_HW} \tag{15}
$$

where Δ*PVRV\_HW\_saved* is the amount of energy saved by the heat wheel in terms of the calculated electric energy consumption of the outdoor unit compared to that without the heat recovery operation in kWh.

The energy saving impact of the heat wheel on the electric energy consumption of the outdoor unit, compared to the system without the heat wheel operation, was calculated with Equation (16):

$$
\Delta P\_{VRV\\_HV\\_s\,\text{sured\\_REL}} = \frac{\Delta P\_{VRV\\_HV\\_s\,\text{sured}}}{P\_{VRV\\_HVO\\_HW}} \cdot 100,\tag{16}
$$

where Δ*PVRV\_HW\_saved\_REL* is the energy saving rate of the heat wheel for the electric energy consumption of the outdoor unit, compared to the system without heat the wheel operation, in %.

The value of the actual energy efficiency ratio of the outdoor unit given obtained the field study was determined with Equation (17) for the investigated cooling season to compare the data provided by the producer:

$$EER\_M = \frac{Q\_{DX\\_HW}}{P\_{VRV\\_HW\\_M}} \cdot 100,\tag{17}$$

where *EERM* is the evaluated energy efficiency ratio (-) based on the measurement during the whole investigated cooling season.

#### *3.3. Formulas for Carbon Dioxode Cross-Contamination in the Heat Wheel*

The scale of carbon dioxide (CO2) cross-contamination in the air-to-air rotary heat recovery wheel was also investigated by measurements in the heat wheel during the operation of the air handling unit in the cooling period. To achieve this, the CO2 concentration difference between the supply inlet and outlet sections of the heat wheel was first determined with Equation (18):

$$
\Delta C\_{\text{CO2\\_cross}} = C\_{\text{CO2\\_HWS}} - C\_{\text{CO2\\_o\\_i}} \tag{18}
$$

where Δ*CCO2\_cross* is the scale of the CO2 cross-contamination in the heat wheel in a given hour in ppm; *CCO2\_HWS* is the CO2 concentration in the supply outlet section of the heat wheel in ppm; and *CCO2\_o* is the CO2 concentration of the ambient air in ppm, which is equal to the CO2 concentration in the supply inlet section of the heat wheel in ppm.

*Energies* **2019**, *12*, 4267

Having completed the measurements, the average of the CO2 cross-contamination values was taken, and the ratio of the result and the supplied average CO2 concentration was calculated by Equation (19):

$$
\overline{\Delta \mathbb{C}}\_{\text{CO2\\_cross\\_AV}} = \frac{\sum\_{i=1}^{n} \Delta \mathbb{C}\_{\text{CO2\\_cross\\_i}}}{n},
\tag{19}
$$

where Δ*CCO2\_cross\_AV* is the average of the CO2 cross-contamination values in ppm and *n* is the number of measurements.

Since CO2 cross-contamination occurs from the exhaust section to the supply section in the heat wheel, the average of the measured CO2 values in the exhaust inlet section of the heat wheel was also calculated with Equation (20) using the data measured:

$$\overline{\mathbb{C}}\_{\text{CO2\\_HWE\\_AV}} = \frac{\sum\_{i=1}^{n} \mathbb{C}\_{\text{CO2\\_HWE\\_i}}}{n},\tag{20}$$

where *CCO2\_HWE\_AV* is the average value of the CO2 concentration in the exhaust inlet section of the heat wheel in ppm and *n* is the number of measurements.

Equation (21) was used to obtain the relative average of differences:

$$
\overline{\Delta \overline{\mathbf{C}}}\_{\text{CO2\\_REL}} = \frac{\overline{\Delta \overline{\mathbf{C}}}\_{\text{CO2\\_rms\\_AV}}}{\overline{\mathbf{C}}\_{\text{CO2\\_HWE\\_AV}}} \cdot 100,\tag{21}
$$

where Δ*CCO2\_REL* is the relative average of CO2 cross-contamination in %, considering the CO2 concentration content in the exhaust inlet section of the heat wheel in ppm.

The maximum value of CO2 cross-contamination was also analyzed during the whole cooling season, which was calculated with Equation (22):

$$\left[ \text{C}\_{\text{CO2\\_REL\\_MAX}} = \left[ \text{MAX} \left( \frac{\text{C}\_{\text{CO2\\_rms\\_j}}}{\text{C}\_{\text{CO2\\_HWE\\_j}}} \dots \frac{\text{C}\_{\text{CO2\\_rms\\_n}}}{\text{C}\_{\text{CO2\\_HWE\\_n}}} \right) \right] \cdot 100 \right] \tag{22}$$

where *CCO2\_REL\_MAX* is the maximum value of CO2 cross-contamination in the heat wheel in the cooling season given by the measured data in %.

## **4. Results and Discussion**

The reference period of the study is the year 2019, more specifically, the cooling period from June 1st to August 31st for a total of 92 days and 25,296 data samples for each of the used measurement points. The AHU is intermittently operated 12 h/day from 8:00 till 20:00 7 days/week. Since this research work focused on the ventilation energy saving of the heat recovery unit's DX cooling coil, the mixing box was shut off during the data recording.

The air handling parameters obtained from the field study for the investigated AHU are illustrated in Figures 5–7 with a monthly timescale. Since the ambient air temperature was the highest in June during the whole cooling season, this relevant month was selected to present the measured data resulting from the data collection.

Figure 5 shows the temperature of the outdoor air (to), the air in the supply outlet sections of the heat wheel (tHWS) and DX coil (tDX), and the exhaust inlet section of the heat wheel (tHWE) over time at hourly intervals in June.

Considering the hottest periods in the cooling season, the ambient air temperature decreased by about 4–5 ◦C due to the pre-cooling effect of the heat wheel, and by an additional 18–20 ◦C, provided by air cooling of the DX coil.

Figure 6 shows the measured relative humidity of the outdoor air (RHo), the air in the supply outlet sections of the heat wheel (RHHWS) and DX coil (RHDX), and the exhaust inlet section of the heat wheel (RHHWE) over time at hourly intervals in June.

**Figure 5.** The air temperature values in the air handling processes.

**Figure 6.** The air relative humidity values in the air handling processes.

The ambient air relative humidity decreased by about 60% due to the air cooling process. In this way, the supplied air relative humidity was around 90%.

Figure 7 shows the enthalpy of the outdoor air (ho), the air in the supply outlet sections of the heat wheel (hHWS) and DX coil (hDX), and the exhaust inlet section of the heat wheel (hHWE) over time at hourly intervals in June.

Considering the hottest periods in the cooling season, the ambient air enthalpy decreased by about 8–10 kJ/kg due to the pre-cooling effect of the heat wheel, and by an additional 30–35 kJ/kg, provided by air cooling of the DX coil.

τ

**Figure 7.** The air enthalpy values in the air handling processes.

Figure 8 shows the sensible effectiveness data (εs) for the outdoor air temperature in June.

**Figure 8.** The sensible effectiveness values as a function of outdoor air temperature in June.

Based on the results, the average sensible effectiveness of the heat wheel was 79.6% during the whole cooling season and the maximum value of 97.6% was recorded in June.

Figure 9 shows the energy saving of the air-to-air rotary heat wheel (QHW\_saved) in terms of the energy consumption of the DX coil, and the cooling energy consumption of the DX coil with the heat wheel operation (QDX\_HW) and without the heat wheel operation (QDX\_WO\_HW), when the DX coil directly cools the hot ambient outdoor air to the supply air conditions during the cooling season.

**Figure 9.** The energy recovery and auxiliary cooling energy consumption for ventilation.

Based on the results, the energy saving of the heat wheel was 2491 kWh in terms of the energy consumption of the DX coil, the cooling energy consumption of the DX coil with the heat wheel operation was 7434 kWh, and that without the heat wheel operation was 9926 kWh.

Figure 10 shows the electric energy consumption of the outdoor unit based on the direct real electric energy consumption measurements (PVRV\_HW\_M) and the calculations made using the recorded air condition parameters with (PVRV\_HW) and without the heat wheel operation (PVRV\_WO\_HW) for the whole cooling period.

**Figure 10.** The electric energy consumption of the outdoor unit.

The real electric energy consumption of the outdoor unit based on the measurements was 1889 kWh and the calculations resulted in 1863 kWh consumption with and 2488 kWh consumption without the heat wheel operation for the whole cooling period.

Since the difference (ΔPVRV\_HW) is only 26 kWh and rate of deviation (ΔPVRV\_HW\_REL) is 1.36% between the values of the measured and calculated electric energy consumption of the variable refrigerant volume (VRV) outdoor unit with the heat wheel operation, Figure 10 shows very good agreemen<sup>t</sup> between the experimental and numerical results. The evaluated energy efficiency ratio is 3.94 based on the measurements (EERM) conducted for the whole investigated cooling season, which is only 0.05 less than the value of 3.99 given by the producer. The energy impact of the heat wheel results in 624 kWh energy being saved (ΔPVRV\_HW\_saved), which is equivalent to a 25.1% energy saving rate (ΔPVRV\_HW\_saved\_REL) in terms of the electric energy consumption of the outdoor unit for the whole cooling period, compared to the system without the heat wheel operation.

Figure 11 shows the measured CO2 concentration of the outdoor air (CCO2\_o), the air in the supply outlet section of the heat wheel (CCO2\_HWS), and the exhaust inlet section of the heat wheel (CCO2\_HWE) over time at hourly intervals in June. There are a few hours in Figure 11 when the recorded CO2 values of the air were lower in the supply outlet section than in the exhaust inlet section of the heat wheel, probably due to the uncertainties and transient response characteristics of the CO2 sensors.

**Figure 11.** The carbon dioxide values in the investigated supply and exhaust sections of the heat wheel.

Having completed the measurements of the whole cooling period, the average CO2 cross-contamination value (Δ*CCO*2\_*cross*\_*AV*) was 63.9 ppm. The average value of the CO2 concentration in the exhaust inlet section of the heat wheel (*CCO*2\_*HWE*\_*AV*) was 390.1 ppm. Based on the results, the relative average of CO2 cross-contamination (Δ*CCO*2\_*REL*) was 16.4% and the maximum value (*CCO*2\_*REL*\_*MAX*) was recorded as 30.1%, considering the whole cooling season. To determine how the obtained values influence the indoor air quality inside of the conditioned spaces, further indoor air quality measurements are necessary (with the use of further measurement devices and questionnaires), which can act as a continuation of this research work, but exceed the limitation of this recent ongoing research project.
