Optimizing Ventilation Systems for Sustainable Office Buildings: Long-Term Monitoring and Environmental Impact Analysis

Abstract

One of the key elements in meeting decarbonisation targets is improving energy efficiency in the building sector. Although much is being done at the policy level, evidence from practice shows that buildings designed and constructed for energy efficiency often do not meet the efficiency targets. This matter has particular relevance when it comes to non-residential buildings, such as offices. A common problem with existing office buildings is the inefficient management of their HVAC systems, which leads to a waste of energy. The goal of this study is to demonstrate, based on the monitoring of four relatively new offices, the extent to which mechanical ventilation leads to energy performance gaps in office buildings and to estimate the resulting environmental impact over the life cycle of the building. The monitored parameters were the occupancy and indoor environment, focusing mainly on the relationship between the actual occupancy and the CO₂ concentration as a parameter representing the performance of the ventilation system. The monitoring results showed that most of the time, the buildings were over-ventilated, with the ventilation rates failing to match the actual demand, resulting in wasted energy. The actual occupancy of the monitored buildings was much lower than their design value. In two buildings, it never reached 50% of the design value. The simulation showed that simply by applying ventilation rate reduction based on a more realistic occupancy schedule, the primary energy demand decreased by 30%. Thus, the building’s annual CO₂ emissions could be reduced by up to 12.5%. These findings help to fill in the knowledge gap as to why the building sector is struggling to decarbonise. The results of this work are of great practical value in showing investors, designers and managers the importance of a properly automated and managed building. The practical value of the results was enhanced by the fact that the timeline of the data covered by the analysis began before and ended after the COVID-19 pandemic, making it possible to assess the fine aspects of managing systems in light of the new realities of a changing work culture and office occupancy.

1. Introduction

The growth of the world’s urban population is driving unprecedented levels of construction, with the area of buildings expected to double by 2060. Today, construction accounts for 39% of global greenhouse gas emissions, including 28% from operations and 11% from building materials and construction. The construction industry has an important role to play in achieving the global sustainability goals of the Paris Climate Agreement [1,2] and net-zero emissions by 2050.

Energy efficiency policies worldwide are driving an increase in the number of energy-efficient buildings, but a considerable savings potential remains untapped. Although buildings are designed to be energy-efficient or nearly-zero-energy, there is still clear evidence that they are not being used as intended. According to the RIBA CIBSE database, buildings typically use 1.5–2.5 times more energy than predicted. This is called a performance gap between the predicted and actual indicators. Gaps in energy efficiency complicate the EU’s efforts to reduce CO₂ emissions, fail to meet the expectations of building owners and are a disappointment to investors.

The causes of performance inconsistency can be identified at all stages of a project [3,4]. The main causes of the energy gap are the following: (1) occupants using more energy than predicted; (2) a larger number of occupants than originally predicted; and (3) problems with energy-saving technologies [5]. Large differences in performance also occur because engineering systems and their components often do not work as intended. Poorly operated and maintained equipment in commercial buildings can waste approximately 30% of the available energy [6,7].

Due to the fact that in energy-efficient buildings, embodied carbon can contribute up to 80% of lifetime GHG emissions [8], several studies [9,10] including by Liang et al. (2019) have suggested focusing more on examining the ratio between embodied and operational carbon to fully explore the decarbonisation potential of an entire building’s life cycle [5].

A literature review also showed that most studies have focused on the improvements in the building envelope [11,12] and left HVAC and automation systems outside the scope of the works [13]. Not much attention is paid to changes in building automation in existing buildings that can sometimes, even without any (or with a relatively small) increases in the embodied energy, lead to a solid operational energy reduction. In a review of 59 papers and 178 case studies, Mirabella et al. (2018) summarized that the automation of buildings was analysed only in 2 percent of all the relevant studies [14]. Dilsiz et al. (2019) also concluded that in existing buildings, operational energy should be primarily reduced in order to reduce greenhouse gas emissions, as it is often obtained only through new construction or retrofitting strategies [15]. One of the possibilities for decreasing operational energy in existing buildings could be by focusing on building HVAC systems, which are great energy consumers in buildings [16,17,18]. Building control/automation systems play a vital role and can be classified as an important element of the whole technical system in buildings [19,20]. Even an improvement in the existing control method can bring improvement up by 23% [21].

It is also important for the future operation of a building to make proper assumptions during the design stage of the building. Bae et al. (2021) paid attention to the fact that during the design stage, designers made assumptions related to HVAC systems, which seemed to be overrated and did not reflect the actual situation [18]. Salimi et al. (2019) and Li et al. (2022) suggested that HVAC sizing could potentially be reduced using actual occupancy information [22,23]. As Harputlugil and de Wilde [24] noticed, “buildings do not use energy: people do” [24].

Today, people spend most of the time in indoor environments; therefore, it is a priority to reach a good balance between energy efficiency in buildings and thermal comfort or indoor air quality (IAQ). The prevalent air pollutant is CO₂, which is often used as an IAQ metric. Therefore, a demand control ventilation based on the occupancy level could be a key solution to optimising ventilation-related energy consumption.

Our literature review revealed that (1) assumptions made during the design stage regarding HVAC system operation and real occupancies often do not correspond to the actual operation of the building and (2) HVAC system automation and management have a significant impact on a building’s life cycle, but in LCA studies, these systems and related embodied emissions are mostly ignored.

Although most studies are limited to single or small groups of buildings and short time intervals, the goal of this study is to perform the long-term monitoring of the occupancy and indoor climate in four office buildings, analyse the results, identify the existence of energy performance gaps (EPGs) and demonstrate the potential impact of the management of ventilation systems on the EPG (or potential energy savings). The measurements performed included periods before the pandemic, total lockdown and post-quarantine periods to demonstrate system operation under changing conditions. The simplified life cycle approach was used to assess the environmental impact of the proposed improvement in systems management over the assumed life cycle of the building, using one of the monitored buildings as an example.

2. Methodology

The methodology of this paper (Figure 1) can be divided into the following stages to assess whether the management of indoor climate systems in buildings is efficient and to identify opportunities to improve energy efficiency and environmental impact: (1) building energy performance analysis; (2) long-term monitoring of CO₂ concentration and occupancy of 4 office buildings; (3) measurement data processing and analysis; (4) building simulation based on real data and modelling of an alternative control strategy for one the monitored buildings; (5) assessment of the decrease in the environmental impacts in terms of CO₂ emissions during the life cycle of the building (operational phase) by changing the ventilation system control strategy.

Building energy performance analysis. At this stage, analysis was performed comparing theoretical and actual data collected for 13 energy-efficient office buildings. Theoretical heating energy consumption was based on the Construction Product Certification Centre’s registration database, where energy performance certificates are registered. The actual energy consumption data were collected directly from building managers. The comparison was possible just in terms of heating energy, as other data were not available and could not be comparable because of changes in the certification methodology. Actual consumption data before comparison were normalised (recalculated to standard design temperatures) as previously described in [25].

Monitoring. The monitoring was performed in 4 office buildings located in Vilnius, Lithuania. All of the buildings were built after 2005 and are certified. All of the analysed the buildings have constant air volume (CAV) ventilation systems with heat recovery.

The monitoring timescale of the buildings included periods before the pandemic, total lockdown and post-quarantine periods. The CO₂ concentrations were measured using the weather station, HOBO MX1102A (measurement range from 0 to 5000 ppm; error ± 50 ppm), while the occupancy measurements were performed with TableAir motion sensors (PIR) (see

Table 1. Monitoring periods and equipment.

Building	Energy Efficiency Label	Year of Construction, Useful Area	Monitoring Period		Sample Data (Hourly Values)
B_1 Engineering company office	B	2017, 2405 m2	From total lockdown to the post-quarantine period	From 5 January 2021 to 27 November 2021	7819
B_2 Engineering company office	B, LEED GOLD	2017, 22,164 m2	Before the pandemic	From 15 July 2019 to 23 December 2019	3862
B_3 Office of R&D companies	D	2008, 6567 m2	Post-quarantine	From 7 April 2022 to 28 April 2022	2653
B_4 University administration offices	B	2014, 4107 m2	Post-quarantine	From 5 May 2022 to 30 July 2022	5381

).

Creating schedules for office occupancy was one of the objectives of the monitoring. TableAir motion sensors were installed in each workplace under desks, with the sensor laser detecting movement and the temperature sensor confirming the presence of an employee. This is how we received the information on if an employee was in their workplace and how much time they spent there.

An example of the layout of the sensors (in building B_1) and the HOBO weather station are shown in Figure 2.

Data processing. The data for the CO₂ concentrations were gathered with 5 min intervals and for occupancy, when a signal was detected. The data were processed and presented at one-hour intervals. Data processing was performed using Excel.

Simulation. Building B_1 was chosen to perform the building simulation, calibration and simulation of an alternative management strategy using the DesignBuilder software. The simulation activities can be divided into the following stages:

I.

Data collection

(1)

Design of data collection (project documentation);

(2)

Actual data collection (measurements, inspections of the installed equipment, consultations with the building facility managers);

(3)

Data analysis (documentation analysis and measurement data analysis).

II.

Theoretical model creation

(1)

Creation of a geometric 3D model based on the project’s documentation;

(2)

Definition of the building’s default occupancy schedules provided in DesignBuilder (as the design schedules were not known, the default schedules of the program were used in light of the intended function of the building and its premises) and setup of the indoor environment design values and default control schedules;

(3)

Development of detailed models of the HVAC systems because simplified models are not suitable for the proper evaluation of building management efficiency.

III.

Model calibration

Model calibration was a very important stage of the simulation. The purpose of calibration was to reduce the gap between the real building’s operation and the model for as much as possible. There is no consensus among scientists about the possible error margins of a calibrated model. Therefore, the calibration process is considered subjective and dependent on the methodology used, the goals pursued and the assumptions adopted [26]. According to Silva et al. (2021), the permissible discrepancy between a model and the actual energy consumption is considered to be up to 20% [27].

In this case, the level of energy consumption for heating was selected as the main indicator for calibration as this value was known from the design stage and was also available from the meters installed in the buildings. During the calibration, the input data of the building were adjusted based on the actual measured monitoring data gathered. When the difference between the actual energy heating consumption of the building and the simulated energy consumption did not exceed 20%, the model was considered calibrated and suitable for the simulation of alternative system control scenarios, more oriented towards real occupancies and demand.

IV.

Simulation of management alternative

At this stage, when real occupancies and real indoor temperatures were included in the model, HVAC system management alternatives, which allowed for the saving of energy through improved control, were simulated. This enabled the estimation of potential energy and CO₂ savings through the operational stage of the building life cycle.

Characteristics of the Analysed Building

Based on the 2D plans of the building, a 3D model was created. The main facade of the building faces east. The DesignBuilder visualization of the created model is presented in Figure 3.

The characteristics of the building’s partitions met the requirements for building partitions of energy performance class B, applied during the construction period. The values of the heat transfer coefficients and air exchange values are based on national norms [28]. The values for the building’s linear thermal bridges for public non-residential C and B energy performance class buildings are 0.2 W/m·K. The standard air exchange value for B energy performance class office buildings is 1.5 1/h at a pressure of 50 Pa.

Also, it was necessary to set indoor air quality parameters and their maintenance regimes, as well as people’s occupancy schedules on weekdays and during different times of the year in the model. Based on the project documentation, the microclimate parameters presented in

Table 2. Design microclimate parameters of the building.

Purpose of the Facility	Temperature in Winter, °C	Temperature in Summer, °C	Supplied Air	Exhausted Air
Administrative	22	24	36 m3/person	36 m3/person
Kitchen	22	24	3 h−1	3 h−1
Rest areas	22	24	3 h−1	3 h−1
Corridors	18	Out of operation	3 h−1	3 h−1
Toilets, showers	22	Out of operation	–	72 m3/h
Technical premises	10–16	Out of operation	1 h−1	1 h−1

were set.

In the theoretical model, ASHRAE occupancy schedules were used (since there were no data on it in the project, the default schedules in the program were selected) and the design quantities of people on the individual premises were applied.

Ventilation systems. The ventilation system PI-1 serves the premises of the 1st floor. The supplied air quantity was 2515 m³/h, the pressure, 280 Pa, the extracted air quantity, 2400 m³/h and the pressure, 250 Pa. Heat recovery is performed by a rotary recuperator, with an air heater power of 7.4 kW. The efficiency of heat recovery was 78.8%. The temperature of the air supplied to the premises was 21 °C. The ventilation system PI-2 serves the premises of the 2nd floor. The supplied air quantity was 4885 m³/h, the pressure, 330 Pa, the extracted air quantity, 4515 m³/h and the pressure, 330 Pa. Heat recovery is performed by a rotary heat exchanger, with an air heater power of 17.4 kW. The efficiency of heat recovery was 76.9%. The air cooler power was 18.4 kW (refrigerant: freon R410A). The temperature of the air supplied to the premises was 22 °C.

Cooling systems. Five variable flow (VRF) systems are installed in the building to meet the building’s cooling needs. OK-1 is designed on the first floor, OK-2, OK-3 and OK-4, on the second floor, and a separate external unit, OK-5, is provided for ventilation unit PI-2. The refrigerant is freon R410A. The total cooling power for the air conditioners was 117.5 kW and 22.5 kW for ventilation. The server room has a separate system with two 2.5 kW power conditioners. The efficiency of the air conditioning systems were an EER of 2.86–3.11 and a COP of 3.41–3.69.

Heating. Heat demand for heating devices was 120 kW and for ventilation air conditioners and reserves, 45 kW. Two condensing natural gas boilers with a capacity of 93 kW each are installed. The boiler efficiency was 97%.

Assumptions for life cycle analysis. A life cycle assessment (LCA) was carried out to demonstrate that small and simple improvements in a building can have a significant environmental impact over the life cycle of the building.

Our research focuses on various management techniques for using the already existing equipment; therefore, the enhancements that are explored do not have a significant impact on the embodied energy and only the operational energy will be further evaluated.

We assumed the following:

-

There is no need for additional embodied energy and associated emissions in order to implement better control of HVAC systems, or this embodied part is insignificant.

-

The impact would be greatest during the operational phase, which is why this is the only phase to be assessed.

-

The life cycle of the building is 50 years.

-

The heat source in the building is natural gas, therefore the CO₂ emission factor based on the national norm [28] was assumed as 0.22 kg CO₂/kWh.

-

The CO₂ emission factor for electricity generation was assumed based on the national norm [28] of 0.42 kg CO₂/kWh.

3. Results

3.1. Building Energy Performance Analysis

A total of 13 Lithuanian buildings that are certified and are relatively new (built within the last 10 years) were analysed to demonstrate the scale of the problem of the energy performance gap [29]. In order to maintain confidentiality, each building is marked as B_1, etc. The monitored buildings have been previously described in detail in Table 1, and they are labelled as B_1 and B_4. A comparison of the actual heating energy consumption with the design values (Figure 4) and EPG (%) was estimated (Figure 5).

It can be seen that these buildings have very large energy consumption gaps ranging from −11 to 601%. Meanwhile, buildings B_1–B_4 that were analysed further have EPGs between 36 and 568%. To sum up, it can be assumed that the predicted energy consumption presented in the building certificate to the owner or user of the building in most cases has nothing to do with reality, and it is necessary to look for both the possibilities of refining the assumptions of the certification methodology and to look for gaps in the operation of the building (system management and maintenance, passive and active behaviors of people, etc.) in the process of monitoring.

3.2. Monitoring Results

Since the results presented above showed that the majority of the analysed buildings were suitable for further monitoring, the buildings whose owners had given their consent and in which it was possible to install and leave relatively safe equipment were monitored.

The monitoring was carried out in four office buildings (see Table 1) in Vilnius (the specific data of the objects are not disclosed, because the building managers did not give permission to make the data public). Since it was possible to monitor only one room in the building with the available amount of equipment, rooms where a larger number of people work (open-type offices) were purposefully selected.

Figure 6 shows the monitoring results of CO₂ and occupancy variation, as well as building occupancy histograms for the four buildings analysed. Separate grey lines present weekly variation in CO₂ concentrations and black bold lines present the average weekly curves from Monday to Friday, with blue lines representing variation in the occupancy.

The monitoring of CO₂ was performed during different periods (Table 1); therefore, some differences can be noted, especially due to the differences in occupancy schedules.

Maximum measured occupancies within the buildings varied in a range of 0.48–0.85 (the highest value was observed in building B_4). In addition, the highest daily peaks were found on Mondays and the lowest, on Fridays. The weekly occupancies gathered are provided in Figure 6.

Figure 7 presents the occupancy frequency and density as persons/sq. m. The occupancy density is a parameter used in some standards and also in simulation tools. When the density is known, the number of occupants can be calculated for different sizes of premises.

According to the national hygienic norms [30], the minimum area for one person should not be less than 6 sq. m (density of 0.167 persons/sq. m), while the maximum recommended area for one person should be 30 sq. m (density of 0.039 persons/sq. m).

Building B_1. Figure 6 shows that the air quality in Building B_1 nearly all the time corresponded to IDA 1 quality (when the outdoor air CO₂ concentration is 400 ppm, the IDA 1 threshold is 800 ppm and IDA 2, 1000 ppm). Only once, when the occupancy peaked at 40%, did the concentration reach the limit of IDA 2 (average quality). The average weekly curve shows that the CO₂ concentration was usually less than 600 ppm. The average occupancy during weekdays (Figure 6) did not exceed 15%. It can also be noticed that the highest average occupancy was on Mondays, but, compared to the other buildings covered by the analysis, the degree of weekly variation was smaller. The building occupancy histogram (Figure 7) proves that almost all of the time when monitoring took place (1917 h), the occupancy density was within the range of 0.033–0.039 person/sq. m, which is the maximum area (about 30 sq. m) for one occupant, as per hygiene norms.

So, in summary, most of the time the room was over-ventilated due to very low occupancies. This was influenced by changes in user behaviours and the switch to remote work even after the quarantine ended.

Building B_2 monitoring was performed before the pandemic. Here, low occupancy rates were observed, although for reasons other than remote work. The office was dedicated to engineering activities, and employees had no assigned workstations, as most of their work was done outside of the office. The maximum occupancy rate in this office was 44% (Figure 6), while the weekly average occupancy did not exceed 28%. Figure 6 shows that the air quality in the building B_2 most of the time was between high and average. The averaged curve shows that the concentration was mostly below 800 ppm.

The building occupancy histogram (Figure 7) shows that the occupancy density range was much wider than in building B_1; nevertheless, for most of the monitoring time it was in a range of 0.033–0.039 person/sq. m, which is the maximum area (about 30 sq. m) for one occupant, as per hygiene norms.

In this building, the energy-saving potential of the ventilation system is lower than in the previous building. However, for the sake of energy saving, switching to IDA 2 can be more economical, taking into account the fact that the same occupant is not in the office all day.

Building B_3 monitoring was performed during the post-quarantine period. In comparison to other buildings, the occupancy rate here was the highest, peaking at 70%, with a maximum average occupancy of 38%, but it was still considered as quite low (Figure 6). The pandemic likely changed the working culture; some of the occupants were working remotely, either fully or to an extent. In this building, we observed higher occupancy rates and CO₂ fluctuations within each working day, with high week-to-week variations (grey curves). Occupancies were higher at the beginning of the week and lowest on Fridays (dropping by more than half), as in other buildings. The indoor air quality was nearly 100% sufficient, with the CO₂ concentration ranging between the highest and average categories. The averaged curves show that CO₂ concentrations were below 675 ppm, so there was still energy waste.

Building B_4 monitoring was performed during the post-quarantine time, and here, we also observed a bit higher of an occupancy rate, which for some days reached 85%, but still, the maximum average occupancy was 30% or lower (Figure 6). The CO₂ concentrations show different curves compared to the other buildings, and higher concentrations were measured even during unoccupied hours—this tendency was observable during the winter time when more employees were working remotely. The concentration peaks at 6:00 p.m. cannot be explained by occupancy, but rather by some changes in ventilation system management. The air quality was nearly always between high and average. It did not exceed 700 ppm according to the weekday averages. The building occupancy histogram (Figure 7) proves that during almost all of the time when monitoring took place (2452 h), the occupancy density was within the range of 0.033–0.039 person/sq. m.

The summarised monitoring results of the CO₂ concentrations for all buildings are presented in

Table 3. CO₂ concentrations in the buildings analysed.

Building	Max. Measured Concentration, ppm	Dominant Concentration, ppm	Working Hours When the Premises Were Over-Ventilated, %
			Did not Exceed 600 ppm	Did not Exceed 800 ppm
B_1	940	501–600	81.7	99.9
B_2	1210	701–800	29.1	90.5
B_3	1203	501–600	66.5	93.0
B_4	1155	501–600	58.0	88.4

.

The results show that in most of the buildings analysed, the predominant CO₂ concentrations did not exceed 600 ppm, and the premises were over-ventilated almost all of the time.

The correlation between the CO₂ concertation and the occupancy rate was also compared to the other indoor parameters measured at the same time and is expressed as the Pearson coefficient of correlation. It represents how the different building parameters measured correlate with each other, with “0” meaning no correlation and “1” and “−1” reflecting a perfect linear positive or negative correlation, respectively. The strongest correlation, compared to other measured indoor parameters, such as indoor temperature, relative humidity, air velocity and solar radiation, was observed between the indoor CO₂ concentration and the occupancy in the room, with the correlation coefficient varying from 0.506 (average correlation) to 0.834 (strong correlation) (Figure 8). This correlation suggests that changes in indoor CO₂ concentration can predict the number of people who are indoors, which can be used in the predictive controls of building HVAC systems.

3.3. Simulation Results

After creating a model that corresponded to the building documentation and performing the simulation, a noticeable discrepancy between the actual and theoretical heat consumption of the energy model was found. During the operation of the building, its management strategy and design microclimate parameters, the schedule of occupancy and the actual number of occupants in the room may have changed. The differences between the design number of people on the premises and the actual number became even more pronounced during the pandemic, when the average building occupancy was just 2 to 23% [28].

The heat consumption of the theoretical model was found to be 11.09% lower than the normalized actual consumption of the building. Following the energetic model development scheme, this model was further calibrated based on the results of the performed measurements. Although a good discrepancy had already been obtained in the case under consideration, it was known from the monitoring phase that the schedules of the internal temperature of the building and the occupancy of people in reality differed from the design ones. Therefore, the model was refined.

The measured average air temperature in the representative office and for the heating system was 23.00 °C. The measured number of people on the premises and the occupancy schedule of the premises were entered in the calibrated model. The monitoring of the premises was carried out during the pandemic, so the load of people on the premises was 33% of the total number of people expected to be there. Using the average hourly data, it was determined that the maximum average number of employees in the room was 5. In the calibrated model, the number of people on the premises was reduced by 67% compared to the theoretical model.

Figure 9 shows that the heat consumption of the calibrated model was 118,938 kWh, and the difference between the actual consumption and the model was 3.85%, so the model is considered to have been calibrated correctly.

As the model was being calibrated, the heat consumption changed just as the electricity consumption for pumps, fans and lighting was reduced. Electricity consumption for cooling was also changed, but cooling could not be properly analysed because of the following reasons: (1) the certificate did not indicate the energy consumption for cooling and (2) the building did not have separate electricity metres for cooling (as was the case for most of the comparable buildings).

3.4. Life Cycle Evaluation of Alternative Ventilation Strategy

After the building model had been created and calibrated, an easily implementable management strategy (one that does not require interventions into systems) was modelled by reducing the amount of air: the model used a reduced amount of supplied air, taking into account the lower amount of people detected during monitoring.

Figure 10 presents the savings using the reduced air flow model. An annual reduction of CO₂ was estimated at 1.44 kgCO₂/sq. m (if energy sources in the building stayed unchanged within the life cycle).

Generally, the operation of a ventilation system is more difficult to evaluate than that of a heating system, because the effective management of the ventilation system is inseparable from the activity schedules of the people in the building. The HVAC system pump energy costs were also slightly reduced with more efficient control, but this amount of energy was negligible in the overall balance. The proposed measures—a reduced amount of air—require no additional investment and no embodied energy.

After applying the primary energy factor for electricity used in Lithuania, which is 2.3, and the CO₂ emission factor for electricity produced in a mixed way, which is 0.42 kg CO₂/kWh (according to [27]), energy savings were expressed through primary energy and the amount of CO₂ saved. It was found that, after applying the strategy of reduced air volume, the primary energy demand decreased from 30 kWh/sq. m to 21 kWh/sq. m, or by 30%.

Meanwhile, after conversion to CO₂ emissions, 12.5% was obtained as the amount of energy that can be saved per year.

When evaluated over the entire life cycle of the building and assuming that the life of the building is 50 years, through the improvement in the management strategy of ventilation alone, it is possible to avoid up to 1.44 kgCO₂/sq. m/year, which is 72 kgCO₂/sq. m during the entire operational phase of the building life cycle, if energy sources stay the same. These emissions can be avoided even without any additional automation equipment, meaning that there will be no additional embodied carbon emissions.

4. Conclusions

COVID-19 and the different periods of the pandemic had a major impact on the office sector, workplace culture, working flexibility and the requirements for HVAC systems. The presented study enabled to the following conclusions:

• Despite the different monitoring periods, the actual occupancy in the four buildings monitored was much lower than the design, never reaching 50% for two of the buildings.
• The CO₂ measurements proved that although the offices that were monitored were quite new, the ventilation systems were not demand-controlled and energy was wasted through over-ventilation. As the ventilation air flow rates were calculated based on maximum occupancies, this led to the fact that for some offices, systems were designed to accommodate excessive air volumes already in the design phase. Flexibility is needed in the design phase, when calculating air flow rates, to take advantage of more realistic occupancies. The problems that were identified obviously prove the need to control and manage HVAC systems better.
• The proposed measure—a reduced amount of air flow rate—requires no additional investments and no additional embodied energy and carbon emissions. It was found that after applying the ventilation of reduced air volume, the primary energy demand decreased from 30 kWh/sq. m to 21 kWh/sq. m, which was by 30%. Meanwhile, after conversion to CO₂ emissions, it was found that an economy of 12.5% could be achieved year-to-year. Assuming that the life cycle of the building is 50 years, it is possible to avoid up to 1.44 kgCO₂/sq. m/year, if the energy source is not changed to another.

The results of this analysis show how important it is to focus on existing buildings and their inefficient management in order to manage climate change and achieve the EU’s decarbonisation targets.

Real energy and environmental efficiency of buildings can be achieved by increasing the level of their intelligent management (automation), using effective management strategies that allow for the synchronization of the levels of buildings more efficiently, addressing user needs and employing smart networks. Broader certification of buildings according to the smart readiness index (SRI), which is currently fairly new and optional, would also help in achieving better management, together with a wider application of the digital twins’ technology.

The main limitation of this study is that there were no BMS data on the ventilation plant operation analysed, as it was not available, but it is planned in future research in buildings with BMS system data to propose optimal management solutions and estimate maximum potential energy savings.