Available via license: CC BY 3.0
Content may be subject to copyright.
IOP Conference Series: Earth and Environmental Science
PAPER • OPEN ACCESS
Evaluation of mitigation effects on air pollutants for electric scooters in
Taiwan with the energy flow analysis and system dynamics approach
To cite this article: P Y Hsieh et al 2018 IOP Conf. Ser.: Earth Environ. Sci. 191 012136
View the article online for updates and enhancements.
This content was downloaded from IP address 158.46.223.200 on 05/11/2018 at 17:17
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd
The 4th International Conference on Water Resource and Environment (WRE 2018)
IOP Conf. Series: Earth and Environmental Science 191 (2018) 012136 IOP Publishing
doi:10.1088/1755-1315/191/1/012136
1
Evaluation of mitigation effects on air pollutants for electric
scooters in Taiwan with the energy flow analysis and system
dynamics approach
P Y Hsieh1,4, L F W Chang1, T Y Yu2 and K C Wu3
1Graduate Institute of Environmental Engineering, National Taiwan University, No. 1,
Sec. 4, Roosevelt Rd., Taipei 10617, Taiwan (R.O.C.)
2Department of Risk Management and Insurance, Ming Chuan University, No. 250,
Sec. 5, Zhong Shan N. Rd., Taipei 11103, Taiwan (R.O.C.)
3Institute of Applied Mechanics, National Taiwan University, No. 1, Sec. 4, Roosevelt
Rd., Taipei 10617, Taiwan (R.O.C.)
E-mail: d00541007@ntu.edu.tw
Abstract. This research establishes a localized dynamic system model to explore changes of
air pollution emission in the transition of electric scooters (ES) considering energy
transformation. The calculation of emission factors (EF) of criteria air pollutants and
greenhouse gases for Heavy-duty Gasoline-powered Scooters (GSH) and Heavy-duty ES
(ESH) is performed with energy flow analysis. Compared with the GSH, the EF of TSP, NOx,
VOCs, and CO2e for ESH reduce by, respectively, 14.8%, 97.4%, 100%, and 76.8% per
kilometer travelled in 2016; although the SOx EF for ESH is 2.4 times higher than that for
GSH, the increment is down to 22.2% in 2025. If the SOx emissions intensity of electricity
reduce to 100 mg/kWh, the SOx EF for ESH will be lower than that for GSH. System
dynamics and energy flow analysis can provide effective analysis about mitigation scenarios
and these findings are helpful to local authorities for air quality management.
1. Introduction
Air pollution caused by the scooters’ emissions is a serious concern in Taiwan with the world’s
highest density of scooters, reaching 378 per square kilometer. To improve the air quality,
Environmental Protection Administration has established and actively promoted a program of “The
development of Electric Scooter” since 1998. However, the market-share of electric scooters (ES) was
not blooming and it held only around an 1% share of the motorcycle market in the past twenty years
even the government offered a subsidy for purchasing ES.
The market share of scooters in Taiwan has changed dramatically in the past three years with the
first Heavy-duty ES (ESH) launched in 2015 (table 1) [1]. The market-share of ES has finally
exceeded 4% in 2017, and exceeded 7% in 2018 to the end of April. Focus on the market for ES, the
relative market-share of Light-duty ES (ESL) fell from 100% in 2014 to 18% in 2017, and its market
share was only 8% from January to April 2018. On the other hand, the market-share of ESH rose
rapidly from 0% in 2014 to 82% in 2017, and exceeded 91% from January to the end of April 2018.
However, the market share of the gasoline-powered scooter (GS) has dropped from 99% in 2014 to
96% in 2017, while it was only 93% from January to the end of April in 2018. In the GS market, the
The 4th International Conference on Water Resource and Environment (WRE 2018)
IOP Conf. Series: Earth and Environmental Science 191 (2018) 012136 IOP Publishing
doi:10.1088/1755-1315/191/1/012136
2
average market-share of Heavy-duty GS (GSH) accounted for 97%, and the one of Light-duty GS
(GSL) accounted only for 0.5%.
Table 1. Market share of Scooter in Taiwan (*the data of 2018 includes Jan. to Apr.).
2012
2013
2014
2015
2016
2017
2018*
Avg.
ES
1.35
1.07
0.76
1.56
2.45
4.41
7.05
2.66
ESH (%)
0.00
0.00
0.00
0.55
1.53
3.61
6.45
1.73
ESL (%)
1.35
1.07
0.76
1.01
0.92
0.80
0.59
0.93
GS
98.65
98.93
99.24
98.44
97.55
95.59
92.95
97.34
GSH (%)
97.15
96.39
96.17
94.77
94.74
92.96
90.73
94.7
GSL (%)
0.50
0.50
0.51
0.51
0.52
0.68
0.23
0.49
Others (%)
1.00
2.04
2.55
3.16
2.29
1.95
1.99
2.14
This study mainly discusses that the changes in the number of heavy-duty scooters with time,
including GSH and ESH, and assesses the reduction of air pollution by substituting a GSH for an ESH
for two reasons. First, the average market share of GSH plus ESH is around 96.05% from 2015 to
2017. The market share of ESH is higher than 92.95%, and that of ESH is higher than 6.45% from
January to the end of April in 2018. Second, ESH in Taiwan has become a more attractive. ESH is also
the battery-swapping electric scooter for now, which improve the disadvantage that wait to charge.
Besides, a company of ESH launched portable battery chargers as an alternative option for customers
to increase the ES’s competitiveness [2].
This study consists of three main parts to gain a more comprehensive assessment of the change of
air pollution emission. Firstly, a new system dynamic (SD) model of the transition to ES with time and
the assessment of localized air pollutant emission of GSH and ESH is built up to understand the
reduction potential for the transitions of ESH and energy. Secondly, calculations of emission factors
(EF) of criteria air pollutants (NOx, VOCs, TSP and SOx, hereinafter CAPs) and greenhouse gases
(CO2, CH4, and N2O, hereinafter GHGs) for GSH and ESH are performed by energy flow analysis,
which the boundary combines associate stationary, mobile and area pollution source. Thirdly, three
scenarios of power structure and different speeds of ESH transitions were set to identify the key
parameters for achieving the reductions of GHSs and air pollution now and in the future.
2. New system dynamic model of the number of electric scooter transition with time
System dynamic (SD) is a useful tool to help address complex issues involving delays, feedbacks and
nonlinearities, and to explore complex long-term policies [3]. In the transportation research, there are
many papers apply SD to study system issues in transportation, such as [4] and [5]. To explore the
growing trend of ES and their influence on the reduction assessment of air pollution under various
scenarios between 2016 and 2035, this paper builds up a new SD model of ESH transition and an
assessment of CAPs and GSGs emission. There are three functions of the model. First, the model
could show the flexible interconnections between the transition to ES and the reduction of energy
consumption and environmental impacts. Second, the model could explore the key parameters for the
development of ESH. Third, the model could assess the possibility that the benefits of carbon and
CAPs reduction in the future.
The main variables influencing the market-share of scooter include the taxes imposed on GSH, the
subsidy for purchasing ESH, the convenience of energy supplements for ESH, etc. Figure 1 shows a
version of ESH and GSH stocks to simulate the dynamic behavior of complex ES transition and CAPs
reduction in this study.
The model lets the tax incentive which is from the GSH and the tax would become the subsidy
which is a part of incentive to buy ESH in substitution for GSH. Another incentive to buy ES is the
ratio of the amount of electricity supplement stations to the gas stations, and this research assumes the
ratio would increase with the growth of ESH. The policy factor in the model that would influence the
The 4th International Conference on Water Resource and Environment (WRE 2018)
IOP Conf. Series: Earth and Environmental Science 191 (2018) 012136 IOP Publishing
doi:10.1088/1755-1315/191/1/012136
3
market penetration of ESH. The market-shares of ESH and GSH are set 4% and 93% in 2017; those
are 97% and 0% in 2035.
Figure 1. Simplified causal loop and stock-and-flow diagrams for GS and ESH stock.
The main variables influencing the emission of scooter include , the number of scooters and
kilometers traveled. The formula for calculating the air pollution emissions of GSH and ESH of the air
pollutant emission is . The indicates the emission (ton/year) in the i-th
type of energy flow path for the k-th pollutant, indicates the respective emission factors for the
k-th pollutant, indicates the average kilometers travelled per year per scooter, and indicates
the respective number of scooters. The next section shows more details of the calculation of .
3. The of CAPs and GHGs for GSH and ESH in Taiwan with energy flow analysis
3.1. The comparisons of between GSH and ESH
Most papers agree that the efficiency of an electric vehicle is higher than a gasoline vehicle and the
emission of GHGs is less considering well-to-wheel, for example [6], [7] and [8], but the reduction of
CAPs would be affected by the power structure of each country.
The which this study built up combine all EFs of components and the energy efficiency in
the energy flow path. The components in the energy flow path of GSH mainly include shipping,
refinery, gas station and final burn in the engine of a vehicle. The components in the energy flow path
of ESH mainly include power sector, transmission and distribution of electricity, charging station and
final stage of ESH. To sum up, the formula of , which combines all emissions in each energy
path stage, is
(1)
where indicates the respective emission factors for the k-th pollutant in the j-th stage of the
i-th path (mg/km), i indicates the type of energy flow path for GSH or ESH, j indicates the different
stages or components in each path, n indicates the number of components in the energy flow path, k
indicates the different air pollutant emission factors in each component, indicates the efficiency of
the t-th stage in the i-th path.
3.2. System boundary for the GHGs and traditional air pollutants emission assessment
In this paper, the boundary, that combines associate stationary, mobile and area pollution source, of
assessment concludes the direct and indirect emission of CAPs and GHGs. The stationary
Market-share
of GSH Market-share
of ESH
Replacement
Total amount of
heavy-duty scooters
Taxes from GS
Incentive to buy ESH in
substitution for GSH
Total amount
of GSH. GSH discard
Scrap rate of GSH
Total amount
of ESH ESH discard
Scrap rate of ESH
+
+
+
+
Ratio of the amount of
electricity supplement
stations to gas station
+ the amount of electricity
supplement station
Scooters Sales
per year The reduction of total
CAP emission (k ton/year)
<avg. annu al traveling
mileages per scooter.>
<NOx EF of GSH
with time> <SOx EF of G
SH with time>
<TSP EF of GSH
with time>
<VOC EF of
GSH with time>
<NOx EF of ESH
with time> <SOx EF of ESH
with time>
<TSP EF of ESH
with time>
<VOC EF of ESH
with time>
fuel tax factor Policy factor.
The 4th International Conference on Water Resource and Environment (WRE 2018)
IOP Conf. Series: Earth and Environmental Science 191 (2018) 012136 IOP Publishing
doi:10.1088/1755-1315/191/1/012136
4
pollution sources include power plants and refineries. The mobile pollution sources include the end
pipes of ESH and GSH. The area pollution sources include gas stations and petroleum shipping.
In Taiwan, more than 81% of gasoline comes from China Petroleum Corporation (CPC) and more
than 75% of electricity comes from Tai-power company in 2016. In this way, this research analyzes
the emission per liter of gasoline production in the refinery of CPC, and the emission per electricity of
Tai-power company. The main data sources in the study are from Taiwan Emission Data System
(TEDS) 9.0 [9], Tai-Power report [10], CSR of CPC [11], AP-42 [12], IPCC Assessment Report [13],
and literatures [14]. The main settings of input parameters in the SD model are below: the intensity of
for electricity production is according to Tai-power system in 2016 [10], and the intensity of
for gasoline production is according to the data of CPC in 2016 [11]; the habit of riders to use
scooters in Taiwan is keeping, and annual sale of scooters will maintain 852,747 which is the average
sale of scooters from 2015 to 2107; the average speed of scooter is 40km/hr. (of GSH is 40.86 km/L
and of ESH is 23.60 km/kWh with the headlights turn on.)
3.3. Scenarios design
Since the EF of ESH is directly related to the structure of the power sector, this study designs the
following three scenarios according to the different source of electricity to figure out the potential of
air pollutant reduction for ES transition.
Scenario 1 (s1): Assuming the electricity is from Tai-power system structure in 2016.
Scenario 2 (s2): Assuming the 20% electricity is from green energy, 30% electricity is from
coal-fired power plants and 50% electricity is from natural gas-fired power plants in 2025.
Scenario 3 (s3): Assuming the EFk of coal-fired power plants, LNG-fired power plants and
oil-fired power plant keep the same as those in 2016, the structure of power sector changes
with the energy transition planned by the government, and the speed of ESH transition
changes with Policy factor.
4. Results and discussion
4.1. The assessment of EFk for ESH and GSH
Under the current energy and power structure in Taiwan (s1), the emission of total CAPs and CO2e are
reduced by 94.3% and 76.8% to substitute a GSH for an ESH. Compared with the GSH, the ESH will
reduce the emissions of TSP, NOx, and VOC by 14.8%, 97.4%, and 100% per kilometer travelled; the
emission of SOx would increase by 136.2%.
However, in s2, the EF of SOx for ESH increase by only 22.2% of the one for GSH, and the EF of
CO2e for ESH is only 18.8% that for GSH. The reduction of EFs of TSP, NOx, and VOC for ESH is
15%, 98%, and 100%. Table 2 and figure 2 show the value and the structure of the EFk in each energy
flow stage for GSH and ESH in s1 and s2. The gray bottom means the main emission source in energy
flow. For GSH, the EFs (mg/km) of TSP, NOx, SOx, VOCs, CO2e in s1 are almost equal to those in
s2. For ESH, the main structure of CAPs distribution is different from GSH, and EFs (mg/km) in s2 of
TSP, NOx, SOx, VOC, CO2e are smaller than those in s2; SOx EF in s2 is only 47.9% that in s1.
Table 2. The EFs (mg/km) and data in each stage of GSH and ESH in s1 and s2 (avg. v=40 km/hr)
TSP
NOx
SOx
VOC
Total CAPs
CO2e
s1
EFk of GSH
80.39
251.44
2.716
1,105.74
1,440.29
65,668
(port to gas station + tail pipe)
(0.39+80.0)
(5.24+246.20)
(2.416+0.3)
(29.24+1076.50)
(37.28+1403.0)
(7,879+57,789)
EFk of ESH
68.52
6.56
6.414
0.26
81.75
15,262
(port to charging station + tail pipe)
(0.52+68.0)
(6.56+0.0)
(6.414+0.0)
(0.26+0.0)
(13.75+68.00)
(1,5262+0.0)
s2
EFk of GSH
80.39
251.34
2.509
1,105.74
1,439.98
65,486
(port to gas station + tail pipe)
(0.39+80.0)
(5.14+246.20)
(2.209+0.3)
(29.24+1076.50)
(36.98+1403.0)
(7,697+57,789)
EFk of ESH
68.31
5.02
3.065
0.03
76.43
12,319
(port to charging station + tail pipe)
(0.31+68.0)
(5.02+0.0)
(3.065+0.0)
(0.03+0.0)
(8.43+68.00)
(1,2319+0.0)
The 4th International Conference on Water Resource and Environment (WRE 2018)
IOP Conf. Series: Earth and Environmental Science 191 (2018) 012136 IOP Publishing
doi:10.1088/1755-1315/191/1/012136
5
Figure 2. Comparison of EFk for GSH and ESH (in s1 and s2).
4.2. The changes of the number of scooters, CAPs, GHGs with time
Figure 3(a) shows the potential reduction of target pollutants in s3 to substitute a GSH for an ESH
with the energy transition and ESH transition from 2017 to 2025. In s3, this study sets the dynamic
transition rate of ESH as the parameter Policy factor varies between 5 and 10. The trend of potential
reduction for total CAPs are getting better. The increment potential for SOx is 145% in 2017 and is
down to only 22% in 2025. The trend of potential reduction for CO2e is from 77% to 81%.
Figure 3. Sensitivity analysis on policy factor and the output variables on the SD model (in s3). (a)
The reduction potential of target pollutants, (b) The yearly market-share of ESH, (c) The number of
ESH (unit), (d) The reduction of total CAPs emission to substitute a GSH for an ESH, (e) The
reduction of CO2e and (f) The total emission of SOx from GSH and ESH.
The annual sale of ESH will be higher than that of GSH within 2022 and 2028 according to the
trend of market share for ESH (figure 3(b)). With a different market share of ESH, the number of ESH
would reach 6 to 9 million in 2035 (figure 3(c)); the reduction of total CAPs and CO2e emission to
substitute a GSH for an ESH would reach 37 to 54 kilotons and 1.5 to 2.1 million tons in 2035 (figures
3(d) and 3(e)). Figure 3(f) presents the increasing trend of SOx emission (174 to 191 tons/year in
80.3 9 80.3 9 68.5 1 68.3 1
251.4 251.3
1,10 5.7 1,105 .7
94.3% 94.7%
70%
80%
90%
100%
-
500
1,000
1,500
GSH(s1) GSH(s2) ESH(s1) ESH(s2)
TSP NOx SOx VOC Reduction of CAPs
EFk of target
CAPs (mg/km)
Reduction rate
(a)
65.67 65.49
15.26 12.32
76.8% 81.2%
60%
70%
80%
90%
-
30
60
90
GSH(s1) GSH(s2) ESH(s1) ESH(s2)
CO2e (g/km) Reduction of CO2e
CO2e EF (g/km)
Re duction rate
(b)
99.98 100.00
97.4 98.0
76.8 81.2
14.8
15.0
-136.2
-22.2
-200
-150
-100
-50
0
0
50
100
150
2016 2017 2018 2019 2020 2021 2 022 2023 2024 2025
VOC NOx CO2e
TSP SOx
Reduction of S Ox (%)
Reductions of VOC,
NOx, CO2e , TSP (%)
(a)
02253
50% 75% 95% 100%
"Market-share of ESH"
100
75
50
25
0
2016 2021 2026 2030 2035
Time (year)
%
(b)
02253
50% 75% 95% 100%
Total amount of ESH
10 M
7.5 M
5 M
2.5 M
0
2016 2021 2026 2030 2035
Time (year)
(c)
K tons/year
v=40
50% 75% 95% 100%
"The reduction of total CAP emission (k ton/year)"
60
45
30
15
0
2016 2021 2026 2030 2035
Time (year)
(d)
M tons/year
v=40
50% 75% 95% 100%
"CO2e Reduction to substitute a GSH for ESH (M ton/pear)"
4
3
2
1
0
2016 2021 2026 2030 2035
Time (year)
(e)
tons/year
v=40
50% 75% 95% 100%
SOx emission from GSH and ESH per year
200
175
150
125
100
2016 2021 2026 2030 2035
Time (year)
(f)
The 4th International Conference on Water Resource and Environment (WRE 2018)
IOP Conf. Series: Earth and Environmental Science 191 (2018) 012136 IOP Publishing
doi:10.1088/1755-1315/191/1/012136
6
2035) from all GSH and ESH. The blue line represents the base case, while the colored bands are the
confidence bands where the data of results can be found with probabilities equal to 50% , 75% ,
95% , and 100% .
4.3. Discussion and suggestion
The emission of total CAPs and CO2e are reduced by 94.3% and 76.8% to substitute a GSH for an
ESH in 2016; however, the SOx emission of ESH is 2.4 times higher than that of GSH according to
the results. The findings will be helpful to provide decision makers with information for analysis. For
example, the SOx emission of ESH would be less than that of GSH if the government applies clean
coal technology or enhances the efficiency in the energy path of ESH to reduce emissions intensity of
SOx from current 236 mg/kWh to 100 mg/kWh and it would be better before 2022.
System dynamics and energy flow analysis can provide dynamic analysis of air pollution reduction
scenarios. The locally-based EFs in this study are helpful in making decision in optimal dispatch for
air emission reduction. The parameters and the model could be further applied to assess portfolios
against multi-criterion objectives such as electric bus and vehicles.
References
[1] Ministry of Transportation and Communications R.O.C Taipei: The number of scooter
registrations (in Chinese). (accessed on 8 April 2018) Available at: https://stat.motc.gov.tw/
[2] Kuo C. Gogoro sales grow amid market downtrend Taipei Times 2018 (accessed on 8 April
2018) Available at: http://www.taipeitimes.com/News/biz/archives/2018/04/06/2003690743
[3] Sterman J 2000 Business Dynamics: Systems Thinking and Modeling for a Complex World
(New York: McGraw-Hill)
[4] Struben J and Sterman J 2008 Transition challenges for alternative fuel vehicle and
transportation systems Environ. Plann. B. 35 1070-97
[5] Matthew G, Nuttall W, Mestel B and Dooley L 2017 A dynamic simulation of low-carbon
policy influences on endogenous electricity demand in an isolated island system Energ.
Policy 109 121-31
[6] Ke W, Zhang S, He X, Wu Y and Hao J 2017 Well-to-wheels energy consumption and
emissions of electric vehicles: Mid-term implications from real-world features and air
pollution control progress Appl. Energ. 188 367-77
[7] Wu Y and Zhang L 2017 Can the development of electric vehicles reduce the emission of air
pollutants and greenhouse gases in developing countries? Transport. Res. D-Tr. E. 51 129-45
[8] Onn C, Mohd N, Yuen C, Loo S, Koting S, Abd Rashid A, et al 2017 Greenhouse gas emissions
associated with electric vehicle charging: The impact of electricity generation mix in a
developing country Transport. Res. D-Tr. E. http://dx.doi.org/10.1016/j.trd.2017.06.018
[9] Inquiry System of Taiwan Air Pollution Emission Teds.epa.gov.tw. 2018 (accessed on 8 April
2018) Available at: https://teds.epa.gov.tw/Default.asp
[10] Tai-Power company 2017 KPI of air pollution emission improvement in Tai-Power thermal
power plant (Taipei: Tai-Power Company)
[11] China Petroleum Corporation 2017 Sustainability report 2017 (Taipei: CPC Corporation)
[12] EPA United States: AP-42: Compilation of Air Emissions Factors (accessed on 8 April 2018)
Available at: https://www.epa.gov/air-emissions-factors-and-quantification/ap-42-compila
tion-air-emissions-factors
[13] The Intergovernmental Panel on Climate Change Fifth Assessment Report World
Meteorological Organization; 2009. Available at: https://www.ipcc.ch/pdf/assessment-report
/ar5/wg1/WG1AR5_Chapter08_FINAL.pdf
[14] Elgowainy A, Han J, Cai H, Wang M, Forman G and DiVita V 2014 Energy Efficiency and
greenhouse gas emission intensity of petroleum products at U.S. refineries Environ. Sci.
Technol. 48 7612-24
02253
50% 75% 95% 100%
"Market-share of ESH"
100
75
50
25
0
2016 2022 2028 2034 2040
Time (year)
02253
50% 75% 95% 100%
"Market-share of ESH"
100
75
50
25
0
2016 2022 2028 2034 2040
Time (year)
02253
50% 75% 95% 100%
"Market-share of ESH"
100
75
50
25
0
2016 2022 2028 2034 2040
Time (year)
02253
50% 75% 95% 100%
"Market-share of ESH"
100
75
50
25
0
2016 2022 2028 2034 2040
Time (year)
