Dynamic Simulation of C onstruction Waste in Macao
Lei Kampeng’, Lu lianggang’, Chan Waikin*
“The University of Macau, Macao
> Macao Tourism and Casino Career Center, Macao
Address : Avenida OuvidorA raga, No.46/48, Edif. Nga Lim, 13 Ander(D), Macao
Tel.: +853-66875146
E-mail address: drali1964@ yahoo.com
Abstract This paper is the analysis for the behavioral tendency of the construction
waste (CW) volume in Macao from 2006 to 2025. Four sources of CW are selected to be
the objects of study, which are assumed to constitute all the CW in Macao. Some related
factors, such as area of Macao, the average stay time of tourists, population density are
also taken into consideration. STELLA 8 is used to perform the analysis, and
correlation analysis of parameters will be carried out by a statistic software SPSS
(SPSS Inc., Chicago, Ill.). The simulation result shows that the total CW will reach
530128 cubic meters in 2025. The sum of total CW from 2006 to 2025 will have a
volume of 13,818,250 cubic meters. From the results of the simulation, the largest
portion of CW is generated by casino and hotel projects, which is the main source of CW
in the entire simulation period.
Keywords: Macao; Construction Waste; Systems Dynamics; Simulation; Stalla; SPSS.
1. Introduction
Macao is strategically located at the Pearl River Delta of the southeastem coast of
Mainland China. With its long association with Portugal, Macao has been playing a
vital role as the cultural and economic platform linking China and the Portuguese
speaking countries (Fig.1).Macao consists of the Macao Peninsula, Taipa Island,
Coloane Island, and some reclaimed land (Cotai). In total, it covers an area of 28.2 km?
(DSEC, 2006).
The construction industry began to thrive in the late 1980s and early 1990s. The
construction industry continued to prosper from that time until 1995. However, from
1995, the construction industry began to slip constantly until 2002 when the
liberalization of gambling industry was realized and the regional economic situation
became stable again. With such positive factors, the construction industry in Macao
managed to rebound from the seven-year-long down trend.
In 2002, the construction industry began to soar largely due to the liberalization of
gambling industry. CW includes the construction and demolition wastes (abbreviated as
CW behind), it was positively proportional to the development of a city. In recent years,
the construction industry is developing by leaps and bounds due to the casinos and hotel
projects in Macao, in contrast to 2000 and 2001 when the construction industry slipped.
The casinos and hotel projects make the construction industry in Macao bloom, and
therefore construction sites are seen everywhere when traveling in Macao. Besides, the
infrastructure, such as roads, car parks, public parks and other social facilities must be
built in order to satisfy the overall development of the city. As a result, the waste
generated during construction is increasing isochronously. Since most of the CW can not
be incinerated, it has to be dumped in designated areas of Macao. For environmental
concem, the volume of the CW are taking up more and more spaces and new places for
dumping CW need to be found since the current dumping areas will soon reach their
capacity limits.
Fig. 1 The location of Macao.
Hence, the cause of CW and the prediction of it are of great significance to
environmental protection and urban planning. The Macao government needs to foresee
this environmental problem in order to tackle it in advance (The Environment Council,
2006). System dynamics is a powerful methodology and computer simulation modeling
technique for framing, understanding, and discussing complex issues and problems.
According to the data published by DSEC of Macao government, the volume of CW in
2003 began to soar due to the commencement of some large-scale casino projects.
According to the table published by DSEC (2006), the amount of CW in 2005 is 1294863
m?, which is over 2 times that quantity of 2004. The yearly increasing rate of CW is 122
% (table1). With this dramatic increasing rate, we believe that the CW is speedily posing
an environmental problem to Macao.
Table 1 the quantity of different kinds of solid waste in Macao from 2000 to 2005
Solid wastes(ton) 2000 2001 2002 2003 2004 2005
Domestic(ton) 138290 138111 146642 154067 154527 162131
Commercial & industrial 41 307 41 254 46 308 46 019 51508 55 456
CW (m’) 296860 217252 244930 349090 583380 1294863
Waste collected in sea (m*) 1914 2472 4188 3 678 3.894 3612
Most traditional statistical forecasting models, such as the geometry average method,
saturation curve method, least-squares regression method, and the curve extension
method, are designed based on the configuration of semi-empirical mathematical
models. The structure of these models is simply an expression of cause-effect or an
illustration of trend extension in order to verify the inherent systematic features that are
recognized as related to the observed database (Dyson, 2005). Stella has previous been
succeeded to simulate the dynamical trend of Macao’ population (Lei and Wang, 2006),
tap water consumptions (Lei et al., 2006) and the reclaim projections (Lei and Wang,
2008) since 2004. Case study of municipal solid wastes with Stella can be found in
previous study in Texas (Dyson and Chang, 2005), Dhaka (Bala and Sufian, 2007).
Furthermore, Hsiao et al. (2002) had employed it to forecast the change of CW in
Taiwan, here our research presents the trends of CW generation associated with four
different subsystems models using a system dynamics simulation with the software -
Stella 8.
2. Methodology
Modeling has been used for years to help scientists and policy makers to find solutions
to complex problems. It is one of the most valuable and useful applications of
mathematics. Modeling and simulation are intellectually creative and quantitatively
rigorous, and it is a ways of connecting ideas with reality. To simplify the explanation,
we definite the nomenclature of the CW in Macao as (Fig. 2):
CWCH: CW generated by construction of new casinos and hotels
CWIR: CW generated by interior renovation of residence
CWCP: CW generated by Civil infrastructures and private construction project
CWRO: CW generated by reconstruction of old zones in Macao
cw
CWOH cue cwep cwRO
Tourists
Tourists Population
| Population
Population
y Reconstruction
ear ;
Area Ratio
Vn Renovation
on ——| Ratio cpp
Fig. 2 Components and its correlative parameters in Simulation the CW for Macao
2.1 Dynamic Simulation Tool of CW
Correlation analysis will be performed between various parameters in the relevant
ecological system in Macao, such as CW, population, GDP, tourists, wastes, time. As for
those correlation coefficients which is bigger than 0.9, it is assumed that the level of
correlation is high and therefore reliable. All the statistic data are from DSEC. The
following equation shows the relations in the generation of CW.
CW=f{(P opuoation,GD P,Tourist, Wastes, time,...)= > (CWCH,CWIR,CWC P,CWRO)
We let the CW be the dependant variable, and set up the corresponding regression
equation by using statistics software. A system dynamics simulation tool - STELLA will
be used to perform the simulation, and correlation analysis between the parameters in
the selecting the suitable equations will be carried out by statistic analysis software -
SPSS (SPSS Inc., Chicago, Ill.). Most of the information about Macao CW, especially
parameter values, is obtained from The Statistics and Census Service (2006).
Based on the results of the correlations between the cited data, we obtained some
parameters which are run out by SPSS from the foundational statistical analysis, a set of
the potential formula were listed. Then we selected out the most suitable equations
through the comparison of the real data and the simulated data. Finally, by linked up the
relative equations, the model for simulating the entire volume of Macao CW is formed,
and a graph for the generation of Macao CW from 2006 to 2025 is derived by running
the model.
Sensitivity tests are conducted to ensure the dynamic model to be stable and
responsible, so that the simulation results will not change considerably upon high
uncertainty of parameter values. Important parameters including tourist normal growth
rate, city area in 2025, maximum value for population density and the money spent to
generate 1m’ CW, are chosen to perform sensitivity tests. The sensitivity tests results
are satisfactory and do not show any substantial change of the basic pattems.
2.2 The Modeling Process
Steps of modeling in this research are summarized as Fig.3 (Andrew, 1999). In this
research, the model comprises mainly five parts as showed below: Population, CWCH,
CWIR, CWCP and CWRO (Fig.4). The STELLA software (Costanza and Gottlieb,1998
(a); 1998 (b); Odum and Odum, 2000) offers an opportunity to create dynamic visual
models for studying a wide variety of problems. We used a dynamic model created in
STELLA and regression analysis performed using the SPSS statistical software (SPSS
Inc., Chicago, IL) to simulate the emergy trends for Macao. For this analysis, we used
curve and random methods regression. The suitable regression equations were selected
according to two criteria: 1) the equation had to be logical (i.e., it had to provide a
causal explanation for an observed trend), and 2) it had to have a high goodness of fit
(Le., R’). Details of the modeling process of are 5 subsystems were shown in the
following section 3.
Complexity
(—_ Bounding the Selection _)
Problem in time, —
Space,Subsystems
Data Quality of
Requirement “| available Data
1___.
Conceptual
Diagram
Equation
Verification
Sensitivity
analysis
Calibration
(oa
Validation
Je Ow
—
Policy Maker |
Fig. 3 Flow chart of modeling process of the CW for Macao
3. Simulation of the Subsystems of CW for Macao
For calculating the total CW, we will connect the five sub-systems together to
aggregate the generation of CW in Macao from 2006 to 2025. The integrated model is
shown in Fig.4. The five sub-systems are put closely and connected by connectors,
Population subsystem was the foundational block of the all other subsystems. The 4
components of the CW were connected to form the total CW. Total CW of Macao was
the summing result of the four subsystems, i. e, CWCH,CWIR,CWCP,CWRO.
3.1 System Dynamics Modeling of Macao population
The population is the fundamental element of a city’s growth. It has great effect on the
development of a city. Population growth is a complicated dynamic process, which is
determined by various factors. In this way, we find out the key factors that have close
relation with the growth of population.
These factors are the birth rate, death rate, natural growth rate and immigration. We
use STELLA to set up a dynamic model to simulate the dynamic change, including the
statistic data of birth rate, death rate and immigration in 1983 to 2007. We assume that
the birth rate will fluctuate between the rate of 0.0075 and 0.009; death rate will range
from 0.003 to 0.0025. As for the immigrants, we set 3335 people for 2005, which are
obtained from the official website. And we assume the immigrants will range between
2000 and 4000 people a year from 2006 due to the outstanding labor importing policy
from the authorities, when Macao face the human resource lack when the tourism
industry was boomed in the past five years (DSEC, 2006). We also assume that these
parameters were random in the simulating processes.
E=}s) Sector 2: CWCH as wor Sector 1:Population ae
tourist normal
fe growth rate ith vate eat rate
year mui wen Immigrant
tourist acty
growth ra
bith death
yearly touristincrease population
normal increase rate
land increase Ss Sector 3: CWIR za
re ey average tute CF) :
rate multiplier
umber of
residences
cwiR
CWIR storage
CW generated renavation rate
population per residence for esidencee
Sally Population (A embne
mutiptier
actual land indvease safe
yearly land increas
city area year
oy Sector dCWCP ae ‘mato
omy. Sector 5:CWRO Aaa
-yeary/expenditure for civil recontructio,
infrastructure Lost yearly AW of private tale
generate 1m3.CW consyyiction projects
‘CWRO Stoage
CW generates
per capita for th year
CW generated
investment per capita for 3rd year
for private construction CW generated CW generated
project per capita for Ist year Pet Capita for 2nd year
‘cop Yer
Fig. 4 Model components and its correlative parameters in Simulation the CW for Macao
The simulation equations adopted in the population subsystem was listed as:
Sector 1: Population
population(t) = population(t - dt) + (birth + Immigrant - death) x dt
INIT population = 488.144
INFLOWS:
birth =birth_ratexpopulation
Immigrant = if time=2005 then 2.0755 else if time=2006 then 2.6658 else if
time=2007 then 2.17 else if time>2007 then random (2, 4) else 0
OUTFLOWS:
death = death_ratexpopulation
birth_rate = random(0.0075,0.009)
death_rate = random(0.003,0.0025)
The simulated result was shown in Fig. 4.
&® 2 population 2: birth 3: death 4: Immigrant
1 8509
7
2
30
a: 450
2 4
3: 1
4 0 -
2005.00 2010.25 2015.50 2020.75 2026.00
Page 1 Years 06:16 F 4 2008447) 1841
a af id Population
Fig. 5 Simulation graphs of Macao population from 2006 to 2025
In the next 20 years, the births and deaths in Macao will decrease but we expect it to
be stable by the fact that better economy will help to ease the burden of raising children.
The population of immigration will range from 2000 to 4000 people a year to prevent
excessive population expansion while emigration can be neglected for the number is
small to the increase of population (Fig.5). In 2025, the total population of Macao will
be 608500.
3.2. CW generated by casino and hotel projects (CWC H)
The main force for building new casinos and hotel are the gambling company who
want to provide more gambling table and equipment to the tourists, so as to get more
profit. In fact, with the liberalization of gambling industry, the quantity of CW will
surely increase to some extent. Before finding out the results, necessary information is
collected to perform the analysis. As for the CWCH, some reasonable assumption
should be made since the Macao govemment only provides information about the
overall CW quantities from 2001 to 2005.
In the first place, we try to find out the tendency of CW growth if the liberalization of
casino industry were not carried out. We believe that in such a case the CW in Macao
7
will only grow in a moderate way. Now, we take a look in the CW generation in 2001 to
2005 in the neighboring area, like Hong Kong, as a reference. The data that we obtain
in table 2 are from the environmental bureau of Hong Kong (EPD, 2006).
Table 2 CW in Hong Kong (2001-2005)
year 2001 2002 2003 2004 2005
CW(x1000 ton) 6408 10202 6728 6595 6556
From the data, we notice that the generation of CW in Hong Kong remained steady
from 2001 to 2005 except for the year 2002. Since Macao was very similar in the city
development style, so we may guess that the construction industry in Macao would be
performing like Hong Kong without the liberalization of gambling industry.
In this way, we assume that the quantity of CW in Macao from 2002 to 2005 will
remain more or less the same as in 2001 if the liberalization of gambling industry were
not carried out. According to the survey of the local construction company, we therefore
assume that the CW will increase with a yearly increase rate of 1.5%, and 2.0%, 5.0%
and 6.0% for the year from 2002 to 2005 respectively excluding the CWCH. Then we
can calculate the CWCH, the calculation results are shown in Table 3.
Table 3 Estimated CW in Macao
Year 2001 2002 2003 2004 2005
Estimated CW in Macao without
new casinos and hotels
Actual CW in Macao 217252 345881 349090 583380 1294863
* Estimated CWCH 0 125370 124169 347213 = 1044526
217252 220511 224921 236167 250337
* Estimated CWCH =Actual CW in Macao - Estimated CW in Macao without new casinos and hotels
In order to perform the correlation analysis between the CWCH and the tourists, data
of tourists from 2001 to 2005 is collected (see Table 4). In this way, we are going to use
SPSS to find out if tourist has reliable correlation with CWCH in recent years.
Table 4 Number of Tourists in Macao from 2001-2005
Year 2001 2002 2003 2004 2005
Tourist (x1000) 10279.0 11530.8 11887.9 16672.6 18711.2
The equation is CW=3270806.49-532.487+0.022 Tourists”
here R? was 0.929, which can be considered reliable. For carrying out the simulation
later, we are going to find out the normal growth rate of tourists in Macao.
The equation was:
Tourists =15531.8 + 2123.695sequence - 138.802sequence”
While “sequence” referred the sequence numbers of relative year till the final 20. In
this equation the R? was 1, we will use the yearly growth rate of tourist in the simulation
latter. The above equation was used to simulate the CVCH.
In our assumption of the intervals of construction projects from 2005 to 2025, the
principle we use is to make the intervals of construction projects continuous according to
the current situation of construction industry for casinos and hotels. Generally, the
construction period of a new casino and hotel is about 3 years in Macao, according to the
survey of the ongoing projects data, he assumption of the construction projects interval are
shown in Table 5.
Table 5 The ongoing Casino and Hotel Project from 2001 To 2012
2001 2002 2003 2004 2005 2006 2007 2008 2009
-2004 -2005 -2006 -2007 -2008 -2009 -2010 -2011 -2012
Number 1 1 3 6 3 2 1 1 1
Period
The final project is assumed to be scheduled to commence in 2009 and end in 2012,
when all the casinos and hotels in the Cotai Gaming strip are supposed to be completed.
New projects of casino and hotel will become less and less, and the CWCH will be
mostly renovation works. Since the newly built casinos and hotels occupy large area,
we can foresee that their renovation works will also generate a considerable amount of
CW. From 2012, we have added a component which will generate a certain amount of CW
each year. This is due to the renovation works in casinos and hotels. The first periods from
2012 to 2018 will generate 80000 to 100000 cubic meters of CW, while from 2019 to 2025
there will have 90000 to 120000 cubic meters of CW. The reason for setting two different
amount of CW is that more casinos and hotel will have to do some changes for the interior
environment and renovation several years after their completion. For this, we can generate
a curve based on the previous assumption which indicates the relation of CW from one year
to another (Fig. 6)
We will use this curve to adjust the CWCH in this chapter. The method to adjust the CW
is to multiply the original CW with the multiplier. The multiplier is obtained by dividing
the quantity of CW each year by the quantity in 2006. We use these multipliers to simulate
the CWCH. The simulation equations adopted in the CWCH subsystem was listed as:
Sector 2: CWCH
area_of Macao(t) =area_of Macao(t- dt) +(yearly_land_increase) x dt
INIT area_of_Macao = 27.5
INFLOWS:
yearly_land_increase = area_of_Macaoxactual_land_increase_rate
tourists(t) = tourists(t - dt) + (yearly_tourist_increase) x dt
INIT tourists = 18711
INFLOWS:
yearly_tourist_increase = touristsxtourist_actual__growth_rate
actual_land_increase_rate =land increase__rate_multiplierxnormal_increase_rate
CWCH = (3270806.5-532.487 xtourists+0.022 xtouristsxtourists) xmultiplier
daily_average_tourist = tourists/365xstay_time_mutiplier
daily_Population = population+daily_average_tourist
normal _increase_rate = 0.03
population density = daily Population/area_of_ Macao
tourist_actual__growth_rate =
tourist normal__ growth _ratextourist_increase_rate_multplier
year = time
land_increase__rate_ multiplier =GRAPH(area_of_Macao)
(27.5, 1.00), (28.1, 0.9), (28.8, 0.8), (29.4, 0.7), (30.1, 0.6), (30.7, 0.5), (31.4, 0.4), (32.0,
0.3), (32.7, 0.2), (33.3, 0.1), (34.0, 0.00)
multiplier = GRA PH(year)
(2005, 1.00), (2006, 0.815), (2007, 0.453), (2008, 0.299), (2009, 0.23), (2010, 0.103),
(2011, 0.0322), (2012, 0.0394), (2013, 0.03), (2014, 0.0373), (2016, 0.0292), (2017,
0.0376), (2018, 0.0506), (2019, 0.0517), (2020, 0.0517), (2021, 0.0414), (2022, 0.05),
(2023, 0.0487), (2024, 0.0474), (2025, 0.062)
stay_time_mutiplier = GRA PH(year)
(2005, 1.50), (2007, 1.54), (2009, 1.59), (2011, 1.64), (2013, 1.69), (2015, 1.74), (2017,
1.78), (2019, 1.84), (2021, 1.89), (2023, 1.94), (2025, 2.00)
tourist_increase_rate_multplier = GRAPH(population_density)
(19.8, 0.995), (20.3, 0.97), (20.8, 0.81), (21.2, 0.535), (21.7, 0.39), (22.2, 0.27), (22.7,
0.165), (23.1, 0.08), (23.6, 0.04), (24.1, 0.01), (24.6, 0.00)
tourist normal__ growth rate = GRAPH(year)
(2005, 0.122), (2006, 0.143), (2007, 0.139), (2008, 0.134), (2009, 0.129), (2010,
0.123), (2011, 0.118), (2012, 0.113), (2013, 0.108), (2014, 0.104), (2015, 0.0995), (2016,
0.0955), (2017, 0.0917), (2018, 0.0882), (2019, 0.0848), (2020, 0.0817), (2021, 0.0788),
(2022, 0.0761), (2023, 0.0735), (2024, 0.0711), (2025, 0.0688)
&® v0
1 2000000
1 10000004
iF ee
1 0
2001.00 2007.00 2013.00 2019.00 2025.00
Page 1 Years 12:23 F4* 2008¢F 412971
aaF ? Construction Waste Assumpation
10
Fig. 6 The assumption of the CWCH curve
For this, we can generate a curve based on the previous assumption which indicates the
relation of CWCH, and the brief result shown in Fig.7.
B® 2 tourist actual growth rate 2: waste generated by casinos a 3: tourist normal _growth rate
1 om
2 2000000
3 0
1 0
2 10000005}-%
3 0
| oh
L 0
2 0 HE
3: i)
2005.00 2010.25 2015.50 2020.75 2026.00
Page 1 Years 07:21 FF 2008444) 1841
aaF ? Construction waste generated by casinos and hotel projects
Fig. 7 Simulation results of the CWCH
3.3. CW generated by interior renovation in residences (CWIR)
This subsystem is to simulate the generation of CW from the redecoration or
renovation of local residences. We suppose that two out of ten families will carry out
interior redecoration every 5 years. In Macao from the statistical data that most of the
family lives in the apartment with the area of about 600 square feet apartment that is
composed of one parlor, two rooms, one toilet and one kitchen ( Fig.8). And we assume
that the family will renew the floor tiles in the parlor, the wall tiles in the kitchen and
toilet.
Fig.8 A partment unit for calculating the wastes from demolition of old buildings
11
Table 6 Calculated of CWIR
Components width (m) height (m) length(m) volume (m’)
Floor finishing(screeding and tiles) 7.00 0.04 8.00 2.24
Ceiling(plastering and paint) 7.00 0.01 8.00 0.56
Plastering, tiles and paint 0.02 3.00 72.60 4.36
Door,window, other indoor facilities a = + 2.00
Total 9.16 m*
From Table 6, we get the result that the unit volume of CW of the apartment is about
9.16m’, For simplicity, we are going to use 10.0 m’ in the following model simulation.
Now we are able to calculate the yearly CW generated by the renovation works.
We make an assumption that six people constitute a family in average, and then in
2005 there are 81357 residences in Macao of the above apartment unit. Then the yearly
CWIR in 2005 should be: 81357x0.2x(1/5) x10 =32543 m?
The CWIR will increase with the growth of Macao’s population yearly. The
simulation equations adopted in the CWIR subsystem listed as:
Sector 3: CWIR
CWIR_storage(t) =CWIR_storage(t - dt) + (CWIR) x dt
INIT CWIR_storage = 150
INFLOWS:
CWIR =number of__residencesxrenovation_rate xCW_generated__per_ residence
CW_generated__per_residence = 10
number of __ residences = int(populationx1000/6)
renovation_rate_for residences = 0.04
The renovation rate for residences is the factor 2/10 times 1/5=0.04. The result is an
average rate. The simulation graphs and results are shown in Fig. 9, indicating that
these wastes are almost linearly growing upward with the steady increase of population.
&® 1: renovation waste 2: number of residences
1 900000—
2 130000 LZ
ane
1 450000]
2 105000 |
7 x
; |
2 80000:
2005.00 2010.25 2015.50 2020.75 2026.00
Page 1 Years 12:26 FF 2008474 29H
3 aF id generation of CWIR
Fig. 9 Simulation results for CWIR
12
3.4 CW of Civil infrastructures and private construction project (CWC P)
Infrastructure is crucial to the development of a city. Infrastructure need to be
planned out and set up when a city is steadily developed. In Macao, billions of MOP
were invested to build various kind of civil infrastructure in order to suit the
development of Macao in various aspects. Since the money spent for civil infrastructure
is paid by Macao government, we try to find out the trend of the expenditure for civil
infrastructure.
As the government income has close relation with Macao GDP (Unit: MOP, Macao’s
currency, 1 USD=8.023 MOP), so we try to establish a correlation between expenditure
for civil infrastructure and Macao GDP. The GDP and Civil Infrastructure expenditure
Value from 2000 to 2005 are listed in Table 7.
Table 7 GDP and Civil Infrastructure Expenditure Values from 2000 to 2005
year GDP of Macao (x1000) MOP Civil Infrastructure expenditure (x1000)MOP
2000 48972396 865067
2001 49704405 989804
2002 54818745 1344531
2003 63566339 2357179
2004 82899311 3386110
2005 92590984 4331432
* The Civil Infrastructure expenditure is obtained from “GOVERNMENT INVESTMENT AND DEVELOPMENT PLAN”
By using SPSS, the expenditure of civil infrastructure is found to have close
correlation with Macao’s GDP. The R Square value is 0.99 and we get a regression
curve that shows the civil infrastructure grows almost linearly with the growth of GDP.
The equation for calculation is:
Expenditure for infrastructure= -4317249.792+0.123xGDP-3.29E-010x GDP?
With the growing number of tourists in Macao, more tourists will visit the casinos
and the casino income will increase accordingly, and the tax income of Macao
government will increase too. As a result, GDP is affected by the number of tourists in
Macao. In order to predict the Macao’s GDP by the number of tourists, we carry out
regression analysis between these two variables.
We use SPSS to perform the regression analysis between GDP and tourists and get
the results.The R square coefficient is 0.809 which is near 0.81 and it can be considered
reliable.The equation for calculation is GD P= 13552063.828+4079.982xTourists
13
For private projects, the investments can be calculated by the data found in the DSEC
website. In the DSEC website, we have data of the area of completed private building
from 1991 to 2005. We can estimate the yearly investment for private projects by
multiplying the area of completed building by the unit price per square meter. Since the
data of the unit prices can be obtained from 2002 to 2005, we can only work out the
reference unit price in this period. The relevant data are listed in Table 8 and Table 10.
Table 8 Total area of completed buildings and transaction price from 2002 to 2005
Total area of Average priceof Average price of office
Year completed residential units per m? units . Average price of 5
—_ " ¥ industrial units per m’
building (m° ) per m’
2002 102549 6261 10759 2199
2003 243023 6377 9536 2082
2004 215108 7984 10227 2410
2005 391487 10024 13609 3347
Table 9 Area completed of various kind of units
Year Area of completed Area of completed Area of completed Area of other
residential units(m”) office units(m?) industrial units(m’) completed units (m’)
2002 36387 4380 4851 56931
2003 153712 24709 14319 50283
2004 122125 18860 5187 68936
2005 161015 23395 0 207078
For calculating the general unit price per square meter, we first calculate the
proportion of different kind of unit in the total completed area; the results are shown in
Table 10.
Table 10 Proportion of completed units
Proportion of Proportion of Proportion of Proportion of other
year completed residential © completed office completed industrial completed
units/transaction price units/transaction price units/transaction price units/transaction price
2002 0.3548/6261 0.0427/10759 0.0473/2199 0.4448/6406
2003 0.6325/6377 0.1017/9536 0.0589/2082 0.2069/5998
2004 0.5677/7984 0.0877/10227 0.2411/2410 0.1035/6874
2005 0.4113/10024 0.060/13609 0/3347 0.5287/8993
The transaction price of other completed unit is assumed to be the average value of
the preceding three transaction prices. From Table 10, we are able to sum up the four
transaction prices each year with their proportion, and we can get the general price per
square meter from 2005 to 2002 is 9694, 6722, 6367, 5634 MOP respectively.
For the private investments after 2005, we assume that the investments keep
14
increasing until 2010, then it goes down gradually from 2011 to 2025.For the
simulation of CW of this kind, we use the money spent to simulate the CW. Firstly, we
calculate how much money spent to generate one cubic meter of CW. Then we divide
the sum of expenditure of the infrastructure and private projects by the CW of this kind.
In the simulation, the value of general unit price for private building in 2005 (9694
MOP) will be used in the entire modeling process. The simulation result is shown in
Fig. 10.The simulation equations adopted in the CWCP subsystem listed as:
Sector 4:CWCP
CWCP storage(t) =CWCP._storage(t - dt) +(CWCP) x dt
INIT CWCP._ storage = 221630000
INFLOWS:
CWCP=yearly_CW_of_civil__infrastructure_projects +
yearly_CW_of_ private construction_projects
Cost_to__generate 1m3_CW = 32462
GDP = 4079.982 xtourists+13552063.828
yearly_CW_of _civil__infrastructure_projects =
yearly expenditure for civil infrastructure / Cost_to__generate_1m* CW
yearly_CW_of_private_construction_projects =
investment for private construction project/Cost_to__ generate 1m? CW
yearly_expenditure for_civil__ infrastructure =
(0.123xG DP-(3.29E-10)xGD PxGDP-4317249.792)x1000
investment_for_private_construction_project = GRA PH (year)
(2005, 3.8e+009), (2007, 4.2e+009), (2009, 4.4e+009), (2011, 4.4e+009), (2013,
4.4e+009), (2015, 4e+009), (2017, 3.3e+009), (2019, 2.9e+009), (2021, 2.5e+009),
(2023, 2.3e+009), (2025, 2.3e+009)
B® 1: waste generated by infrastru 2: Waste generated by civil infr 3: Waste generated by private ¢
1: 3100004
rs 175000
3: 145000
a 275000 |
2: 1500004
b 105000 |
1 240000 RY,
2 125000 |/ SS;
3 65000— + —
2005.00 2010.25 2015.50 2020.75 2026.00
Page 1 Years 0721 F 4 200844 F1 1811
aaF ? Waste generated by infrastructure and private residential projects
Fig. 10 Simulation graphs result for CWIP
3.5 CW generated by reconstruction of old zones in Macao (CWRO)
CWRO in Macao locates in the old district in the North of Macao peninsula. The
15
CWRO has also a large quantity. We assume that the speed of reconstruction of old
buildings depends on the growth of total population of Macao in that almost all the old
buildings in the old district are not more than 5 storey high, which can not accommodate
as much population as high buildings can do. As a result, reconstruction of old
buildings can solve the dangers from deterioration of building structures, and contain
more people in the newly high buildings. In this way, we assume that the rate of
reconstruction is expressed in terms of percentage in population. In fact, the CW
generated from reconstruction is not a steady quantity, it varies dramatically. For
example, in the first stage of reconstruction is to demolish the structure of old buildings.
In this stage, large volumes of concrete and brick debris will be generated and carted
away. The volume of CW in this stage is much higher than the latter stages. For this
reason, we assume that the construction time for reconstruction of one building needs 4
years, and the CW generated in all the latter stages is about 50% of the volume in the
demolition stage.
Thus, for simple simulation, the quantity of CW generated will be expressed in cubic
meter per capita yearly. We assume that in the old building for reconstruction, a
department of 500 square ft. can accommodate 6 people, and the CW generated from
reconstructing the apartment will be the volume of the unit as shown in Fig. 7.
This residential unit is composed of one parlor, one kitchen, one toilet and two rooms.
We apply the general dimension of reinforced concrete building structure and
architectural finishing to make assumption of the above residential unit. The data are
assumed in Table 11.
Table 11 Assumption of configuration of the existing apartment units to be demolished
Architectural or structural configuration Dimension
Storey height 3.0m
Thickness of floor slab 120mm
Extemal brick wall 200mm.
Internal brick wall 150mm
Primary beam 300X 600mm
Secondary beam 250X 400mm
Column 500X 500mm
Average thickness of floor finishing 30mm
Average thickness of wall finishing 20mm
Then we use the above data to calculate the total volume of the CW of the above
residential unit (Table 12) .
Table 12 Calculation of CW for one apartment unit
Components width (m) height(m) length(m) volume (m)
0.30 0.60 31.00 5.58
Beam
0.25 0.40 8.30 0.83
Column 0.50 0.50 18.00 4.50
Floor slab 7.00 0.12 8.00 6.72
16
0.20 3.00 24.10 14.46
Brick Wall
0.15 3.00 12.20 5.49
Floor finishing
i . 7.00 0.04 8.00 2.24
(screeding and tiles)
Ceiling(plastering and paint) 7.00 0.01 8.00 0.56
Plastering, tiles and paint 0.02 3.00 72.60 4.36
Door, window, other indoor facilities -- + -- 5.00
Total 49.74 m*
The total volume is 49.74m’. However, we have to introduce the density factor into
the calculated result because the density of the CW when dumping is smaller than its
original form. We assume that the total volume has to be multiplied by a density
coefficient 1.3. Then the volume will become 63.93m° .
Since there are other public structure components in the old building, like the corridor,
stairs and roof house, we will multiply the results by another correction factor 1.3 to
compensate for these structure components and make the result reasonable. Then we
divide the result by 6 people, and we are able to get the CW generated per capita. The
value is 13.85 m?, and we will use for 14 m* simplicity.
In this subsystem, we have added a factor to dynamically simulate the CW. The factor
is the population. When the population grows, it demands more spaces to take up the
population. When the space is limited, one solution is to replace those low and old
buildings with some higher buildings which is in line with the general city planning. In
this model, we assume the reconstruction rate be 0.2% of the population. It means that
0.2% of the population, or approximately 1000 people will be involved in the
reconstruction projects. When the population increases, the number of people involved
in the reconstruction projects will increase also. For the initial number of people
involved in the reconstruction projects, it needs about ten buildings of five-storey high
to accommodate.
The CW generated per year is different from one year to the next. The CW that we
calculated above is only for the demolition stage. The latter stage will also generate a
certain amount of CW, but it will be relatively less than in the demolition stage. In this
model, we assume that the reconstruction of one old zone in Macao will take four years.
The first year will generate the most waste. In this year, the demolition and foundation
works will carry out. In the second year, it will proceed with the framing works, which
will produce less waste. In the third year, the construction enters into the interior
partition and decoration works, which will produce more waste than the previous year.
In the final year, the construction is near its end, and the waste produced in this year
will become less. We assume that the wastes generated in the second and the final year
is the same. And we assume that the total CW generate in the entire reconstruction
period will be the total quantity in demolition stage plus its 50% volume. That means
the CW per capita in the four years’ reconstruction period was 14x1.5=21 m’ in total.
17
Therefore, we assume the CW generated per capita in the reconstruction period are
divided into: 1) 14 m’of first year. 2) 1.8 m’ of the second year. 3) 3.6 m’ of the third year
and 4) 1.6 m’ of the fourth year. We will use the following equations in the simulation of
the CWRO.
Sector 5:CWRO
CWRO_Stoage(t) =CWRO_Stoage(t - dt) + (CWRO) x dt
INIT CWRO_Stoage =0
INFLOWS:
CWRO =if time=2010 or time=2014 or time=2018 or time=2022 then
populationx1000xrecontruction_ratexCW_generated__per_capita_for 1st_year else if
time=2011 or time=2015 or time=2019 or time=2023 then
populationx1000xrecontruction_ratexCW_generated__per capita for 2nd_year else if
time=2012 or time=2016 or time=2020 or time=2024 then
populationx1000xrecontruction_ratexCW_generated__per_capita for 3rd_year else if
time=2013 or time=2017 or time=2021 or time=2025 then
populationx1000xrecontruction_ratexCW_generated__per capita for 4th_year else 0
CW_generated__per_capita for 1st_year=14
CW_generated__per_capita_ for 2nd_year=1.8
CW_generated__per capita for 3rd_year =3.6
CW_generated__per capita for 4th year =1.6
recontruction_rate = 0.002
4 Aggregating CW in Macao
We use STELLA to simulate CW from 2006 to 2025. The results are shown in Table
13. The simulation graph is shown in Fig. 11.Since some of the data that we use in the
model are uncertain to some extent; we will therefore conduct sensitivity tests to prove if
this model is still robust if these data vary.
Table 13 The simulation results of CW in Macao from 2006 to 2025
Year | CWCH CWCP CWIR CWRO Total CW
2006 1386871 263921 34100 0 1684892
2007 1190476 282373 36106 0 1508955
2008 947903 292021 37776 0 1277700
2009 799581 298810 39303 0 1137694
2010 466946 301770 40325 16937 825977
2011 192164 303958 41467 2239 539828
2012 136146 304098 42540 4594 487379
2013 126351 303847 43158 2072 475428
2014 128536 299242 43716 18361 489854
2015 126493 294519 44410 2398 467820
2016 125900 283095 44904 4850 458748
2017 163780 271630 45370 2178 482959
2018 195416 265132 45929 19290 525767
18
2019 198281 258586 46598 2516 505981
2020 188372 253195 47277 5106 493950
2021 165223 247804 47843 2296 463166
2022 190812 244338 48322 20295 503768
2023 186157 240873 48857 2638 478525
2024 184429 240488 49471 5343 479731
2025 237680 240103 49948 2398 530128
9 1 total construction waste
1 20000004
1 10000004
J ee ee
1 0
2005.00 2010.25 2015.50 2020.75 2026.00
Page 1 Years 07:21 FF 2008%F 411 18H
aaF ? Total CW simulation from 2006-2025
Fig. 11 Total CW simulation from 2006-2025
5. Sensitivity Analysis
For such analysis, we choose four parameters, which are assumed to be important to
the results of the simulation, to conduct the sensitivity tests. The model will be rerun by
different values of the following parameters.
a) Tourist normal growth rate
b) City area in 2025
c) Maximum value for population density
d) Money spent to generate 1m? CW
The first parameter is tourist normal growth rate. In the model, we assume that the
growth rate is according to Table 4.14 with an initial growth rate of 14.27% in 2006.
Now we assume that the initial value would be changed to 17%, 20% or 23%, but with
the overall trend of the tourist growth rate unchanged. Because this parameter is a
graphic input, we can not directly use the sensitivity function of STELLA. We will
manually change the value of the parameter in the model. Then we are able to get the
graphs of total CW when the tourist normal growth rate changes (Fig.12), the 4 curve
shows very diffident in quantity, so this parameter was sensitivity.
19
| | 7
| |
‘ al Se ae |
ses sso rh Omran page
aes 2 ge ne) BES % — rerstcw von rasres
rT | |
: ee i 1 ‘00000
a en gees —
¥ c . ‘ | | | |
ET alse mas mee Err Ty ma mars ma
ed Yee OS) FF HU is ost Ft mwah
BEF 2 mews curren sans oatortmalgner 20 BEA 7 revsslcw ton sans inet serramsi gone ote)
Fig. 12 The total CW from 2006-2025
(Initial tourist growth rate set as a)14.27%, b) 17% , c) 20% and d) 23%)
The second parameter is the city area at 2025. We assume it to be 34km? earlier, now
we assume that it is possible to be at 30km?”, 32km”, 36km’ and 38km’, then we conduct
a sensitivity test concerning this parameters. The parameter is also a graphic input, so
we will simply change the value of the parameter in the model and run the simulation to
get the graphs. Then we are able to get the graphs of total CW when Macao’s area
changes (Fig.13), the 4 curve shows very diffident in quantity, so this parameter was
sensitivity.
a) b)
oe eee reer
Se ae |
= ‘Tormo Tatas Talks abs ae ‘hose Talos Ta
Bez Pre watcw tan 2062025 eo aes 042 25) oe ee
b)
20
f vecicisacin aie
f |
ato | 7 momo | | |
|
i | } |
= aS
L + 1 4
7oist0 bas wale Er 6th am
age aad Fon a2 rages
BEF ——tetwolca rm 206205 je of can st 36nd 205) ae ?
Fig. 13 The total CW from 2006-2025 when the area set as a) 30 km?.p) 32 km?,c) 36 km? and d) 38 km?
in 2025.
The third parameter is the population density. In the model, we set the maximum
density at 24.55 when the tourists can not grow any further. Now we set it at 22.0, 27.0
and 30.0. Since the parameter is also a graphic input, we will manually change the
value of the parameter in the model and the get the graphs. Then we are able to get the
graphs of total CW when Macao’s population density changes (Fig.14), the 4 curve
shows very diffident in quantity, so this parameter was sensitivity.
“foe Tabs 7 als Er cae)
1 weal ey mea 42301 ae
BEF 2 re wuicw tom 2620
popu densty at 72.0 BES 2 Peete rene
be
cary Taibas
BEF 2 revs cn an 0
Fig. 14 The total CW from 2006-2025, (Maximum population density set at a).22000,
b).27000 and c).30000 person/km?,
The last parameter that we conduct a sensitivity test is the money spent to generate 1
21
cubic meters of CW in the civil infrastructure and private projects. In the model, we use
the calculated value in 2005 (32462 MOP) to generate one cubic meter of CW. Since
this value varies from 2002 to 2005 due to inflation, now we assume the value will
range from 2000 to 5000 MOP, and then we conduct the sensitivity test by the
sensitivity function in STELLA using incremental values from 2000 to 5000 MOP. The
curves show very diffident in quantity, so this parameter was sensitivity.
The graph of the sensitivity test result is shown in Fig.15.
JD total construction waste: 1-2-3-4-5-. 7-8
1 2000000
a 1000000%4
1
0
2005.00 2010.25 2015.50 2020.75 2026.00
Page 3 Years 05:45 F4 20084412371
Neer ?
Fig. 15 Sensitivity results for the money flow to generate 1 cubic meter of CW
In summary of all above sensitivity test, the basic patterns of the graphs are all alike
despite we use different parameter values to simulate the same model. That means the
model is immune to the uncertain parameter values, and we can say this model is a
strong, or robust model.
6. Conclusion
In the research, system dynamics models are built up for the simulation of the CW
generation in Macao. The results of the simulation show that the CW will be generated
in an increasing way from 2013. In 2010, 2015, 2020 and 2025 the percentage of
different CW is shown in table 14.The results are summarized as follows:
1) The total CW will reach 530128 cubic meters by 2025.
2) The CWCH will reach 237680 cubic meters by 2025.
3) The CWCP will reach 240103 cubic meters by 2025.
4) The CWIR will reach 49948 cubic meters by 2025.
5) The CWRO will reach 2398 cubic meters by 2025.
22
Table 14 The distribution of the simulated CW in 2010, 2015, 2020 and 2025
Percentage in Total CW 2010 2015 2020 2025
CWCH 73.89% 57.75% 72.39% 77.84%
CWCP 22.77% 37.99% 24.20% 19.34%
CWIR 2.35% 4.05% 3.07% 2.69%
CWRO 0.99% 0.22% 0.33% 0.13%
The sum of total CW from 2006 to 2025 is 13,818,250 cubic meters. Its volume is
equal to a cube of 240X240X240 meters.
From the results of the simulation, we know that the largest portion of CW is generated
by casino and hotel projects, which are the main source of CW during the entire
simulation period.
The simulation result indicates that the total CW generated in 2025 will up to around
530128 cubic metres and its peak value is counted in 2009 with a volume of 1,
684892m*, It shows that the increasing volume of CW requires a large dumping area to
accommodate such a large volume of non-incinerated waste.
From the Table 14, we notice that from 2011 to 2016, the proportion of CW from the
civil infrastructure projects and the private projects increases as the construction of new
casinos and hotels is near its saturation.
The CW generated by reconstruction of old zone in Macao only constitute a small
portion of the total CW from 2010, the year which we assume the Macao government
will commence the reconstruction of old zones. In fact, the pace of the reconstruction of
old zone is still pending due to the disagreement between government and the residents
in the old zone concerning the compensation issues and temporarily accommodation
problems. The thriving public and private civil construction has apparently increased the
amount of CW recently. Since CW is generated in great volume, Macao is experiencing
immense pressure on its limited landfill capacity. Long-term solutions are therefore
crucially needed.
Macao government is now facing a high pressure on improving the public
transportation in the entire city, and carrying out reconstruction of old zones. New
projects like extension of Macao airport, Macao overhead railway system project,
Macao-Taipa submarine tunnel will be launched in the near future. The government's
tax income from gambling industry provides a solid base for carrying out these projects.
These projects will last for several years and push the construction industry forward. As
such, we can foresee that the slowdown of building new casinos and hotels will not
cause the quantity of CW to decrease in a significant way.
Acknowledgments
This research was supported by the Science and Technology Development Fund of
Macau (No.022/2007/A2), Macao Special Administration Area, China. The authors
gratefully acknowledge the assistance of Dr. $.L. Huang of the Graduate Institute of
23
Urban Planning, National Taipei University, for his constructive criticism and comments
on an earlier version of this manuscript.
References
Andrew Ford. 1999. Modeling the environment-An introduction to System Dynamics
Models of Environmsntal Systems. Inland Press. Washington, D.C.: United State of
American.
Bala, B.K., Sufian, M.A. 2007.Modeling of urban solid waste management system: The
case of Dhaka city. Waste Management 27: 858-868.
Costanza R, Gottlieb S. Modelling ecological and economic systems with STELLA:
Part II. Ecol Model 1998;112(2-3):81-4.
Costanza R, Gottlieb S. Modelling ecological and economic systems with STELLA:
Part III. Ecol Model 2001;143(1-2):1-7.
DSEC (The Statistics and Census Service), Macao Special A dministration Region. 2006.
Website: http://www.dsec.gov.mo
Dyson Brian, Chang Ni-Bin. 2005. Forecasting municipal solid waste generation in a
fast-growing urban region with system dynamics modeling. Waste Management 25:
669-679
EPD (The Environmental Protection Department), Hong Kong Special Administration
Region . 2007. Construction waste statistic.
Website: www.epd.gov.hk/epd/misc/cdm/trip.htm
Hsiao T.Y., Huang Y.T., YuY.H., Wemick I.K. 2002. Modeling materials flow of waste
concrete from construction and demolition wastes in Taiwan. Resources Policy 28:
39-47,
Lei Kampeng, Wang Zhishi. 2006. The Dynamic Simulation of Macao Population of the
coming 20 years. Journal of Macao Studies. 6(35): 126-129.
Lei Kam Peng, Lam Chi Kin, Wang Zhi Shi. 2006. The Dynamic Simulation of
Population and Water Consumption of Macao. Journal of Macao Polytechnic Institute.
19(4): 60-69.
Lei Kampeng, Wang Zhishi. Emergy Synthesis and Simulation of Macao. 2008.Energy.
33 (4): 613-625.
Odum H. T., Odum E.C. 2000. Modeling for All Scales. Academic Press.
The Environment Council, 2006. The Report of the Environment Condition in 2005.
Top Design Advertising Company, Macao.
24
