A Study of Smart Ventilation System for Maintaining Healthy Living by Optimal Energy Consumption: A case study on Dalarnas Villa

(1)

1

Student Thesis

A Study of Smart Ventilation System for Maintaining

Healthy Living by Optimal Energy Consumption

A case study on Dalarnas Villa

Author: Fasiha Arshad

Supervisor: William Wei Song, Serena Barakat Examiner: Siril Yella

Course Code: H354T

Higher Education credits: 15 credit hours Date of submission: 2021-01-08

At Dalarna University it is possible to publish the student thesis in full text in DiVA. The publishing is open access, which means the work will be freely accessible to read and download on the internet. This will significantly increase the dissemination and visibility of the student thesis.

Open access is becoming the standard route for spreading scientific and academic information on the internet. Dalarna University recommends that both researchers as well as students publish their work open access.

I give my/we give our consent for full text publishing (freely accessible on the internet, open access):

Yes ☒ No ☐

(2)

2 ABSTRACT

Indoor air quality is a measure of clean air with comfort conditions and depiction of lower concentration of air pollutants. It is tedious task to achieve all quality measures at a time with smart energy consumption. This research aims to come up with a solution of how to improve smart ventilation system in order to get clean indoor air with less consumption of electric energy. Many studies showed that scheduled ventilation system has proven to be a good solution to this problem. For this purpose, a long-term sensor data of smart ventilation system Renson healthbox and Luvians data is studied which is operated in Dalarnas villa. This research investigates how this system works in two modes and to improve it by customized scheduling.

A regression model is constructed in which the relationship between airflow and CO2 is shown. For this

purpose, correlation analysis is used in which the connection of bonds between each data features are analyzed. After the feature selection, as a result from correlation matrix, regression analysis is used to find out whether the selected features are linearly related or not. Regression analysis also used for the intent to quantify a model to estimate the flowrate and CO2. A mathematical model is also build to

simulate the flowrate and CO2 with energy consumption.

The results showed that, in order to provide better indoor air quality with efficient energy consumption, a necessary modification of the fan schedule should be done in a way that fan must be started little bit earlier to avoid harmful particles reach their upper threshold limits. This can result in reduction of fan’s maximum speed hence consumption of less energy is achieved.

Keywords: Smart ventilation, Demand control ventilation, Fan energy use, Indoor air quality, Residents living patterns, Indoor CO2, Ventilation modes, Correlation analysis, Regression analysis, Sensor data,

(3)

3

1. Introduction

Ventilation is a process of getting fresh air from outdoor environment and exhausting the polluted air outside. As we know that due to severe weather conditions in winter, people are subjected to stay indoors by keeping their windows and doors air closed. Due to crowded indoor conditions, ventilation systems were introduced to breath fresh even by remaining at home. Maintaining a high level of indoor air (IAQ) has been found to be extremely vital to guarantee the convenience and appropriate health standards for inhabitants of the house. The key cause of the deterioration of air quality in households have now been confirmed to derive from both indoor and outdoor sources, such as gases and/or pollutants, where their health effects have been found to be more harmful under inadequate ventilation, elevated temperatures and high moisture.

It is necessary to understand what indoor pollution is. Many building materials, furnishes, home cleaning products and chemicals in home releases pollutants. Chemicals like volatile organic compound (also known as VOCs) are related to diseases such as respiratory hazards and other various problems. Formaldehyde, which can cause cancer; and endocrine disrupting and chemicals which can lead to cancer from flame retardants, PVC plastics and perfluorinated chemical coatings makes visible health impacts(EWG Home Guide, 2017).In addition, inadequate ventilation and moisture can create a perfect breeding environment for dust mites that can cause allergic reactions, hay fever and worsen the symptom of asthma. Moreover, Radon, a naturally occurring deadly neurotoxin that can percolate through foundation cracks from the ground into basements, may also be a risk, since lung cancer is attributed to exposure. “The American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) has determined that a home's living area should be ventilated at a CFM rate determined by adding 3% of the conditioned space floor area to 7.5 times the number of bedrooms plus one [formula: vent CFM = 0.03A + 7.5 (# bedrooms + 1)] as published by ASHRAE 62.2 in 2013. In a tight home, mechanical ventilation is necessary to achieve this ventilation rate. ASHRAE Standards are revised every three years.” (Energy.Gov, U.S. Department of Energy).

1.1 Factors effecting indoor air quality

Discussion has persisted for several years as to which measure is the most acceptable for indoor air quality. The most frequently used approach, calculating the CO2 produced by humans and released by appliances like boilers and gas cookers and are possibly carbon dioxide (CO2). Humidity and volatile organic compounds (VOCs), which are potential indicators of indoor air quality.

a) CO

2

as air quality

When arguing of global warming, carbon dioxide is also used since it is one of the major greenhouse gases affecting global warming. CO2 is, therefore, a strong measure of the quality of

(5)

5

indoor air in dwellings, where inhabitants and their behaviours are the primary source of emissions, since all humans emit CO2 while breathing and not many other causes. CO2 is occasionally, though, a health concern in itself. Nonetheless, it is a very good measure of human presence and ventilation levels. There is approximately 400 ppm of outdoor air; breathing produces CO2, so the concentration of indoor CO2 will still be at least 400 ppm and typically higher. A 900 ppm indoor CO2 level ensures sufficient air quality, and 1600 ppm suggests bad air quality (CEN, 2019; Successful House Alliance, 2020). In rooms where the need for ventilation is related to the presence of humans, such as in apartments, children’s room, living rooms, dining rooms, CO2 is the most important predictor.

Figure 1. Carbon Dioxide level chart

b) Humidity as air quality

The relative humidity indoors will change in line with the humidity level outdoors on an annual basis. The incidence of house dust mites can be enhanced by a high degree of humidity in the indoor air. The indoor relative humidity should therefore be kept under 45 percent during winter in climates. (Richardson et al. 2005). Generally speaking, to minimize the likelihood of mould formation, high relative humidity levels should be avoided, with negative health problems such as asthma and allergies as a result (Liddament, 1996).

Relative humidity tests have been conducted for several years and are now in industry practice. In places having cold climates, indoor temperatures are normally high during summer and lower during winter. The indoor relative humidity would be somewhat different from summer to winter, with the same ventilation rate in all seasons. In other words, as a measure of indoor air quality, a fixed relative humidity has some drawbacks and is most effective in wet rooms, where the aim is to prevent very high humidity levels.

(6)

6

However, the difference between indoor and outdoor humidity content could be the best measure in terms of absolute humidity, while indoor and outdoor sensors may be needed for this. In this case, the difference of 3.5 g of water vapour per m3_{of air is a fair amount and should be used} year-round to verify if the output of humidity in the home is appropriately matched with the rate of ventilation. (Velux Group)

c) VOC Compounds

According to research’s, volatile organic compounds or VOCs are well known because of producing photochemical oxidants in combination with NOx in air which can cause adverse effects on health. (Yu and Crump 1998). Various artificial construction materials such as particle board (PB) and medium density fibre board (MDF) have been used worldwide in order to erect a house and decorate its interior. Hence, these building materials emit VOCs and formaldehyde which can be toxic to health (Roodman and Lenssen 2002).

There can be a combination of hundreds of chemicals in a single fragrance in a substance, some of which (e.g., limonene, citrus scent) react with ozone in the indoor air to create harmful secondary contaminants, including formaldehyde. Two Researchers have identified 133 distinct VOCs. Limonene, a- and ß-pinene (pine scents), and ethanol and acetone (often used as fragrance carriers) were the most frequently found. Gas chromatography-mass spectrometry was used by Anne Steinmann, a professor of civil and environmental engineering and public relations at the University of Washington, Seattle, and colleagues to analyse VOCs given off by the goods. Twenty-five air fresheners, detergents for washing, cloth softeners, dryer sheets, disinfectants, detergents for pans, all-purpose cleaners, soaps, hand sanitizers, lotions, d-lotions were checked. (Potera, Carol, 2011)

1.2 Research Objectives

This study operates on the basis of a case study of "Dalarnas Villa" to offer suggestions on how we can use sensor data to increase indoor air quality (IAQ) and smart ventilation performance in a small villa.

Because of the constraints of the dataset we have for the moment, just indoor CO2 concentration is used to analyse human presence. As mentioned earlier in section 1.1, that 1,000 ppm indoor CO2 level ensures sufficient air quality, 1,400 ppm in most cases will guarantee decent indoor air quality, and 1,600 ppm suggests bad air quality (CEN, 2019; Successful House Alliance, 2020). Centred on Dalarnas Villa's data, the following research questions are suggested:

1- Through sensor data analysis to find air flow patterns (in terms of ventilation system) and residents’ living patterns (daily life, in terms of CO2, etc.)

(7)

7 The aims of this research, therefore, are:

1- Obtain sensor details on the efficiency of different mechanisms of ventilation in Dalarnas Villa and how the existing modes of ventilation react to the changing the rate of CO2 and assessing the benefits and drawbacks.

2- Suggest how to improve air quality by utilizing optimal energy consumption to eliminate air pollutants like relative humidity, carbon dioxide and VOC compounds.

3- To determine an optimal pattern with lower energy consumption and good air quality. 4- To make a best schedule and use of the ventilation system to achieve a sustainable living

environment, i.e. lowest energy consumption for a best air quality.

1.3. Scope

This study can play a vital role in the future planning and development of smart ventilation systems. With careful analysis and study of long-term sensor data, some issues of current ventilation systems were brought in light because we can still be able to provide better indoor air quality by saving electric energy. Also, maximum recommended concentrations for appropriate levels for suspended particulate matter, indoor air quality is indicated. Volatile organic compounds, particulate matter and CO2 substances about their possible impact on wellbeing. However, this study is limited to specific Dalarnas villa hence the results can lead to further study. The customized fan schedule is represented based on CO2 pattern, its effects on the ventilation and indoor environment.

2. Literature Review

Using many internet resources, a relevant literature on indoor air quality was undertaken. Over the last decade, many researchers have targeted different areas such as schools, banks, offices, residential areas including single, double and multi-dwelling buildings. Not only this, they also tried to minimize the energy consumption. N.L Seeresha examined the relation of carbon dioxide and the temperature and ventilation of building air, with the marker for determining air quality and ventilation efficiency with carbon dioxide. (Sireesha, N, 2017). Lei Zhao suggested that the frequency of high indoor particulate emission cycles could be decreased to some degree by the mechanical ventilation system; however, year-round testing found that it was not successful in increasing safe time ratios. (Zhao.L.L, 2018). Max H. Sherman, Iain S. Walker & Jennifer M introduced strategies for measuring either equal ventilation or equivalent indoor air quality, showing that equivalent ventilation can be used as the basis for complex ventilation control, peak

(8)

8

load reduction and outdoor contaminant infiltration. Gruber, M., Trüschel, A. & Dalenbäck carried out an experiment in which a workplace location was modelled and the occupancy details used for ventilation control by a model-based controller was eventually exposed to different errors and time delays by taking the predicted values as references into account. The findings revealed that, in many situations, the calculations were adequate to achieve high control efficiency, but above a certain level, only the increased sophistication of the controller was able to resolve the deficiencies. The recovered items included systematic reviews, libraries of evidence-based research and studies on IAQ (Indoor air quality). Via the Implementing Agreement on Energy Conservation in Buildings and Community Structures (ECBCS), a systematic analysis of occupant actions with regard to ventilation affecting Belgium, Germany, Switzerland, the Netherlands and the United Kingdom was initially developed under the auspices of the International Energy Agency (IEA). The aim of this study (Annex 8, Dubrul 1988) was to incorporate questionnaires and observations to ascertain what steps are being taken by inhabitants in order to ventilate their homes and to examine the reasons for these actions. In order to expand this area of operation, further IEA and other foreign events have also taken place. IEA programmes such as ECBCS Annex 27 on 'Evaluation and Demonstration of Domestic Ventilation Systems' (Månsson 2001), ECBCS Annex 28, 'Innovative Annex 28,' are significant examples. (Liddament MW. 2001)

An approach is designed on the basis of above mentioned literature review to explain and accomplish the key objective of the task. In the coming chapters, we will address numerous methodologies that have been carried out in the process of designing the solution.

3. Data Description

In this study, we researched the ventilation system in the "Dalarnas Villa" experimental building, which is an ongoing Dalarna University project in Dalarna in collaboration with Fiskarhedenvillan and Dalarnas Försäkringsbolag.

3.1 Experimental setup for ventilation in Dalarnas Villa

In Dalarnas Villa, Renson Healthbox 3.0 has been introduced. It's a smart house idea in which, along with other ecological resources, several structures are illustrated. A significant part of this project is to explore how harm to a house can be minimized due to issues with moisture and water, while studying how to introduce an indoor environment that is ideal for living.

Healthbox 3.0 is part of the C+ framework at Renson. Using the distinct invisivent window airflow, ffresh air is supplied in dry spaces and the dirty air is eliminated in a smart way from wet places. The moisture level, CO2 and/or VOC (volatile organic compounds or 'odours') are constantly tracked by the built-in central sensors. Healthbox automatically adjusts when the air quality in a certain space declines. The centre of this Renson C+ device only emits contaminated air if needed, based on continuous CO2, moisture, and VOC (volatile organic compound) measurements in the indoor air through sensors per room in the building. If Individuals are present in a given space (and a high CO2 content guarantees the exhaled air) and / or moisture or odours are detected (depending on the type of sensors per room), the level of ventilation will be increased automatically in order to conduct the required escape of contaminated air. However, the central ventilation unit works at

(9)

9

the lowest level while no one is at home and/or the moisture level is favourable and there are no odours, and therefore less air is discharged. Renson also works on construction programs and has an external fan with a roof duct that facilitates demand-controlled ventilation. (Renson Healthbox 3.0).

Figure 2: Demand-Control Ventilation System (Renson Healthbox)

The ventilation fan could operate in two separate modes in Dalarnas Villa under the Central Healthbox command: auto or manual mode.

· Auto: The ventilation fan works on a set schedule.

· Manual: Ventilation fan is on demand control depending upon the concentrations of CO2, VOC or humidity.

Why Renson Healthbox? - A brief economic review

A Health box saves energy compared to having a system running on full speed all the time. The Health box is a cost effective way (cheaper) to achieve healthy indoor climate. The payback time is typically 8-10 years for the extra cost. It’s a combination of quality/efficiency of system. The most common alternative would be full mechanical ventilation with heat recovery. Those units are large and expensive and consume more power for their fans, with ongoing filter and maintenance costs as well. To guess a figure, we might imagine slightly higher installation, say €4000 in total, plus annual running costs of €200.

Dalarnas villa is a test bench, and therefore economic considerations were not the largest factor in deciding which systems to install. The Healthbox certainly has the potential to significantly improve indoor ventilation, compared to either natural ventilation or basic exhaust ventilation.

(10)

10

3.2. Zone Description of Dalarnas Villa

The data is obtained via the Healthbox central discharge unit, which is fitted with integrated sensors that continuously monitor the indoor air level of CO2, temperature, humidity and VOC per linked room in the building. Data is logged every five minutes via sensors in each room.

The Renson Healthbox is connected to 7 rooms, which covers all the zones in which individuals perform their everyday activity in the house.

Following is the description of each zone or room of the house. Zone 1: Kitchen and downstairs living room

Zone 2: Upstairs bedroom east Zone 3: Upstairs bedroom central Zone 4: Downstairs bedroom

Zone 5: Downstairs shower room and toilet Zone 6: Downstairs laundry room

Zone 7: Upstairs bathroom and toilet

(11)

11

Figure 4. Blueprint of Dalarnas villa (Upstairs)

3.3. Dataset

This study is a collection of data starting from 2019-08-03 00:00:00 to 2020-06-31 23:00:00 with the logging time of 5 minutes, and the main variables for every zone are mentioned in the table given below.

Table 1. Important features of each zone

Zone  Time, Mode

 Ventilation flow rate

 Ventilation fan power

 Fan pressure  Fan speed  Flow rate  Fan voltage  Temperature… Zone 1,2,3,4  CO2  relative humidity  rh20, temperature  Zone AQI  flow rate …

Zone 5,6,7  relative humidity

 rh20

 temperature

 VOC (concentration)

 Zone AQI

(12)

12

4. Research Methodology

The flow map in figure 5 illustrates how analysis was carried out in this process. It begins with data collection from sensor which are installed in the villa under experimentation, then the data is pre-processed to carry out further analysis. Analysis using collected data in excel 2016 done to look trends and relationships of harmful particles and ventilation flow rate.

Pre-processing steps involves separation of multiple zones which were recorded in single sheet, to study them independently and data is also divided into two ventilation modes which is auto mode and manual mode. In manual mode, fan speed varies with respect to the concentration of CO2 or harmful particles, while in auto mode, fan continues to run with a constant speed independent of indoor air quality. This is further explained in chapter 5. After separating the modes, data visualizations were carried out in order to see pattern of CO2 and ventilation fan. A threshold of CO2 (upper and lower) is determined to evaluate changes in fan speed.

After visualizing the data, the relationship between each data feature is analysed, but there is a need to see mathematically the bond between them is either strong or weak. For this purpose, correlation analysis was conducted to inspect which variables have stronger relationship with each other. It helps us to build the regression model having two variables (dependent and independent). After correlation analysis, the two features (CO2 and flowrate) were selected which could help us to do quantitative analysis by using a model. We need to carry out regression analysis because we want to see how linearly the selected features are. This is really important part of this methodology. This is a well-grounded method of recognizing which features have impact on a topic of concern. The regression approach helps to confidently decide which variables matter most, which variables can be overlooked, and how these variables affect each other.The relationships between a series of independent variables and the dependent variable can be defined by using regression analysis. Regression analysis produces a regression equation where the relationship between each independent variable and the dependent variable is represented by the coefficients. (Frost, 2020)

(13)

13

Figure 5. Flow chart of methodology

4.1 Correlation Analysis

Correlation analysis is the core part of methodology as to build our model, we need to determine two variables which are strongly linearly related to each other. For this purpose, correlation matrix is build. Analysis of correlation is a computational approach which is used to measure the frequency of the relationship between two variables. A high correlation means that there is a strong association between two variables, while a low correlation means the variables are barely related. Two traits (variables) may be compared with one another positively. This assumes that the value of the other variable(s) often changes as the value of one variable increases.

If the coefficient of correlation is closer to value 1, then there is a positive relationship between variable X and variable Y. A positive association means a rise in one factor correlated with a growth in the other. In the other hand, the closer correlation coefficient to -1 would mean that there is a negative relationship that would result in a decline in the other variable 's rise. The correlation is 0 if two variables are independent of each other. (Cornellius, Towards data science).

(14)

14

Figure 6. Correlation analysis

4.2 Regression Analysis

Regression analysis helps us to foresee the situation whether two variables can be a good fit to build a regression model. In this study, simple linear regression analysis has been done after getting results from correlation analysis matrix. In general, the simple regression is the regression approach to discuss the relationship between one dependent variable (y) and one independent variable (x). (Yan, Xin 2009)

Y = β

0

+ β

1

𝑥 + ∈

(1)

In equation (1) Y is the dependent variable, β0 is the Y intercept and β1 is the slope of linear regression line. Also, 𝑥 is the independent variable, and ∈ is the random error. The response variable is also called the dependent variable, and the explanatory or indicator variable is called the independent variable. R-squared, or coefficient of determination, is one of the indices for calculating the consistency of fit. It is the percentage of variance that the better line model describes. This relies on the ratio of the regression model's amount of square error (SSE) and the sum of the difference in squares around the mean. (Teknomo, Kardi, 2015)

4.3 Modelling

Very first step in investigating the relationship between two features is to plot a graph and visualize the linear relationship between them. (UWE, 2020) We can see from the figure 7 that carbon dioxide

(15)

15

and flowrate are strongly related with each other. So, a model is devised in which one variable is independent (CO2) and other one is dependent (flowrate).

Figure 7. Pearson correlation coefficients

According to correlation matrix in figure 7, we can see that “flowrate” and “CO2” has good corre lativity i.e. 0.86. This yields us to make a good estimation on flowrate as independent variable. T he R-squared value of our linear model is 0.9381 which is near to 1. The higher the R-squared va lue is, the better is the model. We can consider linear model to be statistically significant by deter mining the p-values are below than pre-determined significance level of 0.05. (MachineLearning Plus). The t-value can be viewed something like this. A greater t-value means that, simply by cha nce, it is much less likely that the coefficient is not equal to zero. Therefore, the higher the t-valu e is, the better it is.

(16)

16

Figure 8. Regression on CO2(ppm) and flowrate(m3/h)

In above figure, black dots represent the actual data while green colour shows predicted data. The remaining red ones are residuals. The predicted green line shows that model has a very good fit.

5. Analysis and Results

5.1 Identification of current ventilation modes

As we know from chapter 1 that the current system under study is Renson healthbox 3.0 and is embedded in Dalarnas villa. It is previously discussed in chapter 3 that Renson healthbox have two different modes of ventilation. Auto and manual. To identify the behaviour of flowrate in manual mode, we have constructed plots between flow rate and CO2 to determine the thresholds.

(17)

17

Figure 9. Visualization of manual mode in different zones.

From figure 9, it can be seen that when the CO2 concentration is below 800ppm, the flowrate keeps at constant level and when CO2 is above > 950 flowrate is 100% and keeps running at its full capacity until CO2 level drops down. In between 800ppm and 950ppm fan flowrate changes linearly from 30% to 100%. 0 10 20 30 40 0 200 400 600 800 1000 Fl o wr at e CO2

Zone1

0 5 10 15 20 0 200 400 600 800 1000 Fl o wr at e Co2

Zone 2

0 5 10 15 20 0 500 1000 1500 Fl o wr at e CO2

Zone 3

0 5 10 15 20 25 0 500 1000 1500 Fl o wr te CO2

Zone 4

(18)

18

Figure 10. Analysis of flowrate in auto mode.

In auto mode flowrate keeps constant at 30% to 77% as shown in figure 10. We can summarize above theory as follows.

Table 2. Flowrate (m3/h) of different zones

Ventilation mode

CO2

concentration(ppm)

Zone 1 Zone 2 Zone 3 Zone 4

Manual < 800 22% 30% 30% 25%

800 - 950 30-100% 30-100% 30-100% 30-100%

>950 100% 100% 100% 100%

Auto 30 - 75% 30 - 77% 30 - 77% 30 - 77%

5.2 Analysis of harmful particles in air

A combination of solids and liquid droplets floating in the air is particle contamination, also called particulate matter or PM. Such particles are emitted directly from a single source, while others form in the atmosphere by complicated chemical reactions.

Particles come in a vast spectrum of shapes. Particles with a diameter of less than or equal to 10 micrometres are so small that they can penetrate the lungs, possibly causing significant health concerns. Ten micrometres are shorter than a single human hair's diameter.

 PM1 is highly fine particulate matter of less than 1 micron in diameter. These are

considerably less than 2.5 PM. Their scale is 70-100 times less than a human hair's diameter. The lungs do not filter PM1 particles and can pass the blood-brain barrier. (camfil)

 Particles of coarse dust (PM10) are 2.5 to 10 micrometres in diameter. Sources involve

activities for crushing or grinding and dust kicked up on roads by automobiles.

 Fine particles (PM2.5) have a diameter of 2.5 micrometres or less and can only be

observed by an electron microscope. All modes of combustion, including motor engines, power plants, residential wood burning, forest fires, agricultural burning, and some manufacturing processes, emit small particles. (AirNow, 2017). Outdoor PM2.5 level has an impact on ventilation, and it is crucial to get understanding of how particulate matter is carried from outdoors to the indoor environment via ventilation.

(19)

19

Figure 11. PM1, PM2 and PM3 concentrations in a day

Figure 11 demonstrates the graph of PM1, PM2.5 and PM10 concentrations in a typical workday. We see that PM1 reaches 96 while PM2.5 reaches above 150 and PM10 is touching the scale to 200. This could lead to serious problems if not ventilated properly. According to figure of acceptable PM concentrations, humans can experience breathing discomfortness if PM2.5 touches scale to 100 while PM10 can cause similar hazard when it lies between 251-350. PM1 is the most dangerous particle.

(20)

20

Figure 12. Acceptable PM concentrations

5.3 Identification of CO

2

pattern

CO2 pattern helps us to identify the presence of humans in the house. We can make a good estimate that at which time house is under 0 occupancy and at what times family comes home. Since the CO2 pattern of everyday cannot exactly be same but we it can help us to guess the general pattern. Figure 13 is a graph of CO2 which describes that peak is formed at 20:00 indicating that family is in the house and valley is occurring at 15:00 which tells us that house is under 0 occupancy. The pattern tends to increase after 15:00 and keeps on increasing till 20:00. At midnight, CO2 remains between 500-600ppm. After 11:00 CO2 pattern starts declining as family is leaving for their school or work. However, this is one-day example graph we cannot assume same pattern of CO2 each day.

Figure 13. Co2 pattern in a typical work day

Similarly, CO2 pattern on a typical weekend is different from the working day. It is difficult to predict that when family is at home and when the home is under 0 occupancy during weekends. There are many peaks observed in a weekend as shown in figure 14. At midnight, highest CO2 is observed i.e. 900ppm to 1000ppm. While two valleys are formed in the afternoon and evening where CO2 level remains between 400 – 500ppm. It then starts increasing continuously from 15:00 till 20:00 at night. CO2 pattern is unique for each weekend hence we cannot predict what is going on weekends.

(21)

21

Figure 14. Co2 pattern on a typical weekend

5. Analysis of Ventilation fan, airflow and CO2:

Effective execution of any intelligent ventilation technique requires awareness of the relationships between ventilation speeds and concentrations of indoor pollutants, whether with regard to balance connections or temporal reaction. This includes an understanding of how to be exposed to health and IAQ are affected by pollutants of concern.

A ventilation norm is related to reasonable IAQ by several tacit assumptions. These tacit assumptions need to be compensated for and used in establishing an equivalence theory, or equivalent assumptions. One belief is that airflow actually improves IAQ (i.e., that bringing in outdoor air has a net benefit of diluting indoor contaminants rather than a negative effect of bringing in outdoor pollutants). This is reflected in the use of the equation of normal continuity:

C + A.C = S (2)

Where the concentration (C) is related to the rate of air shift (A) and the intensity of the source (S). (Sherman, M.H., Walker, I.S., Logue, J.M., 2012)

In order to observe the ventilation fan speed with respect to the concentration of air pollutants, we have observed some thresholds of CO2 to change the fan speed to maintain healthy indoor environment.

In figure 15 shown below, we can observe some interesting facts which were observed in the time period of 1 randomly selected week:

1- At the start of each day (midnight), CO2 level remains between 850 to 1000ppm because typically, carbon dioxide levels rise during the night when people are sleeping, especially

(22)

22

if the door and windows are closed. The concentrations then fall during the day if the room is unoccupied. Unfortunately, poor air quality can hinder restful sleep and optimum health in many homes (Zehnder. 2015)

2- It also shows that minimum airflow is 30.10 while the zone4 is under 0 occupancy.

3- When CO2 level touches 750ppm, airflow drops to its minimum level i.e. 30.10. it means that minimum threshold of CO2 for the fan to run at its minimum speed is 750ppm. 4- As soon as CO2 level touches 800ppm, airflow starts increasing from its minimum level

showing that that upper threshold of CO2 is 800ppm.

5- As CO2 level extends to 950ppm, airflow level starts increasing resulting in increase of fan speed at its maximum.

Figure 15. Visual representation of carbon dioxide and effects on airflow and flowrate.

By looking at above chart, one might observe that in order to maintain high quality living indoor environment with energy saving system, we need to stop the harmful particles or gases to reach its upper limit which can cause serious health issues. If we look at the graph carefully, we can divide days into two categories; working days and weekdays.

In a typical working day, a family follows almost same routine every day. By the time 7:00 family leaves the home and return home by 18:00 pm in the evening. We can design a schedule in a way that before 18:00, ventilation fan starts working to avoid CO2 reaching its upper threshold. Similarly, we know that CO2 reaches at its highest peaks at midnight so ventilation fan can be started before reaching its upper threshold.

0 20 40 60 80 100 120 140 0 200 400 600 800 1000 1200 00 :0 0: 00 05 :1 0: 00 10 :2 0: 00 15 :3 0: 00 20 :4 0: 00 01 :5 0: 00 07 :0 0: 00 12 :1 0: 00 17 :2 0: 00 22 :3 0: 00 03 :4 0: 00 08 :5 0: 00 14 :0 0: 00 19 :1 0: 00 00 :2 0: 00 05 :3 0: 00 10 :4 0: 00 15 :5 0: 00 21 :0 0: 00 02 :1 0: 00 07 :2 0: 00 12 :3 0: 00 17 :4 0: 00 22 :5 0: 00 04 :0 0: 00 09 :1 0: 00 14 :2 0: 00 19 :3 0: 00 00 :4 0: 00 05 :5 0: 00 11 :0 0: 00 16 :1 0: 00 21 :2 0: 00

Fri Sat Sun Mon Tue Wed Thu

A ir fl o w vs f lo wr at e C O 2 Time of day

Flowrate, Airflow and CO2

(23)

23

From figure 1, we can define the thresholds of CO2. Normal level of CO2 lies between 450 to 700ppm. After 700ppm an acceptable level of CO2 is reached. We can say that whenever CO2 reach 700ppm fan should start running at about flowrate of 50%. Below 450ppm flowrate of 30% can be applied (see table 2). Above 1000ppm flowrate of 70% can be applied. In this way, fan power energy can be saved and harmful particles will not reach to its upper threshold causing discomfortness.

Table 3. Flowrate, CO2 and fan power consumption in manual mode

Zone Flow Rate

(air valve m³/s)

CO2 (ppm)

Min Avg Max Min Avg Max

Zone 1 3.9 21 39 597 819 1073

Zone 2 2 13.6 20 689 890 1055

Zone 3 2 11.9 20 557 848 1000

Zone 4 2.89 28.1 34.68 722 944 1350

Table 4. Flowrate, CO2 and fan power consumption in manual and auto mode collectively.

Zone Flow Rate

(air valve m³/s)

CO2 (ppm)

Voltage (V)

Min Avg Max Min Avg Max Min Max

Zone 1 3.9 16.5 39 181 611 1073

2 10

Zone 2 2 9.2 20 220 668 1195

Zone 3 2 8.9 20 255 658 1163

Zone 4 2.89 18.5 34.68 218 754 1350

6. Theoretical Analysis of CO2 and energy consumption

As studied earlier in chapter 5, flowrate behaves in two different ways depending upon ventilation mode selected. Flowrate keeps at a constant rate in auto mode while it changes in manual mode depending upon the concentration of harmful particles and CO2. The objective of indoor air quality control for energy-efficient villas is to maintain the indoor CO2 concentration in the comfort zone with a minimum amount of energy consumption. (Wang S. J., 2012)

The above scenario is explained in terms of energy consumption further. Some reasonable assumptions must be made in order to conduct this study.

(24)

24

1- The indoor air CO2 is supposed to be mixed all the time.

2- Volumetric ventilation rate in and out of the building to keep incoming and outgoing air at equilibrium.

3- The amount of incoming air should be equal to outgoing air. In other words, the volume of air should be at equilibrium.

Figure 16. Representation of CO2 and energy consumption

As mentioned in figure 16, three normal curves are represented. If we look at red curve it shows the actual situation where when the carbon dioxide reaches upper threshold, the fan will start running at its maximum. From our data and experimentations, we know that fan starts running when CO2 upper threshold reached level 800ppm (as represented in figure 15). It will keep running at its maximum speed until CO2 reaches its minimum threshold 750ppm (as shown in figure 15). Since fanis at its maximum between these levels, hence it is utilizing its maximum energy. To reduce this energy consumption if we start the fan little bit earlier the harmful particles starts getting discharged in outdoor environment earlier and clean air could be achieved.

As shown in figure 16, blue curve shows a decreased amplitude as compared to red curve. We see that volume of CO2 is unchanged but a little reduction in peak. This means that less energy is used while CO2 peak is also reduced when fan is started a little earlier. But when we look at green curve

(25)

25

it represents that fan is started even earlier than blue one. Hence it will run even at lower speed and lower energy will be consumed resulting decrease in amplitude of the curve.

So, in above three cases we find green curve as optimal as compared to the other curves. In order to understand it even better, we can see examples of each case below.

Case 1:

Red curve (demonstrated in figure 16) defines the actual data. According to experimentation, family wakes up at 8:00 am and after having breakfast and getting ready, they leave home. Near 18:00 family comes back home and CO2 level start rising until 20:00. When it reaches 850ppm fan runs at its highest and consumes its highest amount of energy until threshold starts decreasing and touch 750ppm lower threshold. Near 23:59 in zone 4 Co2 level rise till 1300ppm as occupants are sleeping. During these hours, fan starts running at its maximum speed until it drops down. We don’t want CO2 to rise more than 850ppm as it suggests bad air quality and also fan runs at its highest flowrate.

Case2:

We need to consider a theoretical case which is represented in blue curve in figure 16. If we start the fan before family comes home the fan gets started and starts discharging polluted air outside instead of waiting to reach some high threshold. In this case we see that the amplitude of the curve has decreased meaning that level of CO2 is not alarming but average. Also, the fan will not run at very high speed.

Case3:

We need to consider even another optimal case where fan is started even earlier than the curve discussed previously which is represented as green curve in figure 16. In this case, sensors detect that air quality is getting bad and as soon as they sense CO2 reaches between 400 and 500ppm. In this case, fan will run at a very low speed while discharging the polluted air outside hence consuming a very little energy.

We can notice that the volume of these curves remains same but amplitude is decreased.

When to start the fan?

From figure 1 we can see that acceptable CO2 level is till 1000ppm. After it, harmful co2 level starts in which humans can feel uncomfortable or drowsiness. According to our findings (from figure 15) the fan starts when sensors detect 800ppm CO2 concentration. If we start the fan the little earlier at when CO2 concentration is at 500ppm, we can achieve two benefits:

1- The polluted air will get discharged while maintaining a good and comfortable breathing environment.

(26)

26

Comparison of existing and suggested ventilation schedule

Auto mode:

In auto mode, fan flows at constant speed. It does not effect by the concentration on the harmful particles. Even it flows at its constant speed when the house is under zero occupancy.

In this case, more energy is consumed while good air quality is maintained.

Manual mode:

In manual mode, the flowrate of fan depends on the concentration of harmful particles. When the concentration of CO2 is reached 800ppm, fan starts flowing at its maximum speed until it drains down to 750ppm. When the house is under zero occupancy fan flows at its minimum constant speed. We have seen that almost at midnight the CO2 concentration reached 1300ppm and during this amount of time fan flows at its maximum speed hence more energy is utilized.

Customized mode:

In customized mode, when sensors detect the level of CO2 particles are at 500ppm, fan gets started and starts running at an average speed hence its starts discharging polluted air outside by avoiding reaching CO2 at its upper threshold. In this way a lot of energy can be saved.

We see that we can adjust the starting time of fan and it is realistic to change threshold values for CO2.

7. Conclusion and Discussion

This research aimed to study the ventilation system based on Resnson Healthbox and Luvian’s data embedded in Dalarnas villas by collecting the long-term sensor data. We aimed to study the pros and cons of the current ventilation system and try to improve the weak areas. We deeply analysed the CO2 patterns and energy consumption of the system under study. In addition, the patterns of harmful particles in combination with airflow, flowrate and CO2 were identified. Two different datasets have been studied together (Renson Healthbox 3.0 and Luvian’s data) in similar time zone i.e. local Swedish time to monitor the behaviour of the ventilation system depending upon the concentration of harmful particles.

First of all, to answer the research question how to make a best schedule and use of the ventilation system to achieve a sustainable living environment, i.e. lowest energy consumption for a best air quality, we did some data visualizations to see trends and patterns of the data. These graphs helped us to look at data patterns of the ventilation modes. Data is also categorized into various zones in order to better look into each area of the house and the ventilation patterns depending upon the indoor air quality conditions. Next, we did correlation analysis in order to determine the strong relationship between two variables from dataset. After that, regression analysis was carried out on two features as a result from correlation matrix constructed in first step. (Darlington, Richard B.,

(27)

27

and Andrew F. Hayes, 2016). In the end, model verification carried out and we predicted the values based on some inputs.

As a conclusion, in order to save energy while maintaining indoor air quality, it is essential to avoid fan running at its maximum speed. Because our experimentation showed that CO2 reaches 1300ppm during midnight which can create uncomfortable breathing environment. In addition to this, during this high rise in carbon emissions, fan keeps running at its maximum speed which consumes more energy. We can improve this scenario if we start fan a little bit earlier so that ventilation process starts before co2 reach at its upper threshold limits. As described in chapter 6, we can start ventilation when carbon emissions reach 500ppm. In this way we can maintain good and healthy air while saving electric energy.

However, there are few limitations in this study:

1- We cannot study harmful particles in detail due to limitations in dataset. We compared both datasets by putting them in same time zone hence derived results based on similarities in datasets.

2- The customized fan schedule is represented based on CO2 pattern. However, its effects on the ventilation and indoor environment is not validated.

Therefore, more detailed study need to be carried out in order to improve its efficiency. 1- Collect more detailed data based on harmful particles and its effects on air quality. 2- Future work may consist a detailed study on remaining zones (5,6,7).

3- Moreover, this study can be further expanded to check how much energy can be saved during summer and winter.

4- Efficiency of the current ventilation system can also be measured by calculating after how long (time) the concentration of harmful particles are decreased.

All in all, this study is providing a method of improving indoor air quality by smart energy consumption.

(28)

28

References

:

Home Guid, EWG’s Healthy living, https://www.ewg.org/healthyhomeguide, accessed in Sep 2020

Energy.Gov, U.S. Department of Energy, Ventilation. Available at energy.gov/energysaver/ventilation, accessed in Sep 2020

C+ ventilation system. (n.d.). Retrieved from Renson:

https://www.renson.eu/gdgb/products/ventilation/mechanical-ventilation/c-ventilation- System, 2020

Richardson, G., Eick, S., Jones, R. (2005) How is the indoor environment related to asthma?: literature review, Journal of Advanced Nursing, vol. 52, no. 3, pp. 328-339

Liddament, M. W. (1996) A guide to energy-efficient ventilation, AIVC

Månsson L-G, Svennberg, S.A, “Demand control ventilation, a summary of Annex 18”, Document TSR-02-1997, ESSU

Potera, Carol, 2011 Environmental Health Perspectives; Research Triangle Par: A16. Sireesha, N. (1 Jan 2017). Correlation amongst Indoor Air Quality, Ventilation and Carbon Dioxide. Journal of Scientific Research. 9. 179. 10.3329/jsr.v9i2.31107.

Yu C, Crump D (1998) A review of the emission of VOCs from polymeric material used in buildings. Build and Environment 13(6): pp. 357–374, ELSEVIER

Gruber, M., Trüschel, A., & Dalenbäck, J.-O. (2014). CO2 sensors for occupancy estimations: Potential in building automation applications. Energy and Buildings, pp. 84, 548-556.

Max H. Sherman, Iain S. Walker & Jennifer M. Logue (2012) Equivalence in ventilation and indoor air quality, HVAC&R Research, 18:4, 760-773, DOI: 10.1080/10789669.2012.667038 Zhao, L. L. (2018). Impact of various ventilation modes on IAQ and energy consumption in Chinese dwellings: First long-term monitoring study in Tianjin, China. Building and

Environment, 143, 99-106.ELSVIER

Roodman DM, Lenssen N (2002) A building revolution: how ecology and health concerns are transforming construction, Worldwatch Institute

CEN (2019) EN 16798-1: Energy performance of buildings - Ventilation for buildings -Part 1: Indoor environmental input parameters for design and assessment of energy performance of buildings addressing indoor air quality, thermal environment, lighting and acoustics

Velux Ventilation systems. Velux. Velux Group. https://www.velux.com/what-we-_do/research-and-knowledge/deic-basic-book/ventilation/indoor-air-quality. Accessed on Jan 2021

(29)

29

Cornellius Yudha Wijaya, Towards Data Science. https://towardsdatascience.com/what-it-takes- to-be-correlated-ce41ad0d8d7f, accessed in Jan 2021

Frost, Jim. 2020 https://statisticsbyjim.com/regression/when-use-regression-analysis/#, accessed in: Jan 2021

Yan, Xin, and Xiaogang Su. Linear Regression Analysis: Theory and Computing, World Scientific Publishing Company, 2009. ProQuest Ebook Central,

http://ebookcentral.proquest.com/lib/dalarna/detail.action?docID=477274. Created from dalarna on 2020-12-14 14:31:14.

Teknomo, Kardi (2015) Regression model using Microsoft Excel.

https://people.revoledu.com/kardi/tutorial/Regression/GoodnessOfFit.html, accessed in: Jan 2021 AirNow, 2017. Home of the U.S. Air Quality Index

https://cfpub.epa.gov/airnow/index.cfm?action=aqibasics.particle , accessed in Dec 2020 UWE BRISTOL 2020, University of the west of England,

Data Analysis, http://learntech.uwe.ac.uk/da/default.aspx?pageid=1442 , accessed in DEC 2020 Machine learning plus, https://www.machinelearningplus.com/machine-learning/complete- introduction-linear-regression r/#:~:text=Complete%20Introduction%20to%20Linear%20Regression%20i n%20R&text=Linear%20regression%20is%20used%20to,the%20predictor%20variables%20(Xs ). Accessed in Jan 2021 Camfil,https://www.camfil.com/en/insights/standard-and-regulations /pm1-is-most- harmful#:~:text=PM1%20is%20most%20harmful&text=But%20to%20provide%20a%20truly, are%20prey%20to%20PM1. Accessed in Jan 2021

Wang, Z. L. (2012). Indoor air quality control for energy-efficient buildings using CO 2 predictive model. IEEE 10th International Conference on Industrial

Informatics (pp. 133-138). IEEE.

Zhao, L. L. (2018). Impact of various ventilation modes on IAQ and energy consumption in Chinese dwellings: First long-term monitoring study in Tianjin, China. Building and Environment, 143, 99-106

Darlington, Richard B., and Andrew F. Hayes. Regression Analysis and Linear Models: Concepts, Applications, and Implementation, Guilford Publications, 2016.ProQuest Ebook Central, https://ebookcentral.proquest.com/lib/dalarna/detail.action?

(30)

30

Figure References:

Figure 1 Axiomet, https://axiomet.eu/gb/en/page/1954/air-quality-monitoring-indoors/

Figure 2 Renson Healthbox

https://www.renson.eu/gdgb/products/ventilation/mechanical-ventilation/c-ventilation-

System

Figure 3, 4 Zhu Yurong , A Study of Smart Ventilation System to Balance Indoor Air Quality and Energy Consumption

Figure 12 Airveda, https://www.airveda.com/blog/Understanding-Particulate-Matter-and-Its-Associated-Health-Impact

A Study of Smart Ventilation System for Maintaining Healthy Living by Optimal Energy Consumption: A case study on Dalarnas Villa

Student Thesis