emissions trading in hong kong and the pearl river delta region

Emissions trading in Hong Kong and the Pearl RiverDelta region—A modeling approach to tradedecisions in Hong Kong’s electricity industry

Richard C.M. Yam, W.H. Leung *

Department of Systems Engineering and Engineering Management, City University of Hong Kong, Hong Kong

2

e n v i r o n m e n t a l s c i e n c e & p o l i c y 3 1 ( 2 0 1 3 ) 1 – 1 2

a r t i c l e i n f o

Article history:

Received 18 October 2011

Received in revised form

5 March 2013

Accepted 19 March 2013

Published on line 23 April 2013

Keywords:

Air pollution

Regional emissions trading

Emission control

Modeling

a b s t r a c t

In 2002, the Hong Kong government and the Guangdong provincial government agreed to

reduce emissions of sulfur dioxide, nitrogen oxides, respirable suspended particulates, and

volatile organic compounds by 40%, 20%, 55%, and 55%, respectively. There was strong

public demand for the power stations in Hong Kong to reduce emissions. Emission caps

were introduced, with allowances for the trading of emission credits. However, local power

stations were using equipment built in the 1980s and 1990s, making it difficult for them to

meet the new emissions requirements. The situation presented a new challenge, which

involved a choice of either improving the existing equipment, or using emissions trading to

meet the emission caps. This study reviews the background on emissions in Hong Kong and

the surrounding regions, the ‘‘cap and trade’’ system, and the technologies used for power

generation and emission reduction. A modeling approach is adopted to simulate the

equipment, the electricity dispatching requirements, and the costs of either reducing

emissions or trading emission credits. Data from a power station in Hong Kong was chosen

for the simulation. Different options were simulated in the model to identify the optimal

strategy. The results were then compared with the plan for emission reduction. This study

demonstrates that a modeling approach using linear programming can analyze the com-

plicated options involving emission reduction and investments to achieve an optimized

business solution.

# 2013 Elsevier Ltd. All rights reserved.

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/envsci

1. Introduction

Air pollution in Hong Kong (HK) has worsened since the early

2000s. The number of hours of reduced visibility increased

from 480 h in 1997 to 1490 h in 2003 (HKO, 2007). Hong Kong

faces two major air pollution problems: street-level and

region-wide air pollution. In 2002, the Hong Kong government

commissioned a study of the air quality in the surrounding

Pearl River Delta region (PRD). The PRD comprises an area of

* Corresponding author. Tel.: +852 69057108.E-mail addresses: [email protected], [email protected]

1462-9011/$ – see front matter # 2013 Elsevier Ltd. All rights reservedhttp://dx.doi.org/10.1016/j.envsci.2013.03.010

42,794 km with a population of 38.7 million around the

estuary of the Pearl River, whereas Hong Kong comprises only

1000 km2 and a population of 7 million. The report, entitled

‘‘Final Report – Study of Air Quality in the Pearl River Delta

Region’’ (CH2M HILL, 2002), revealed that the main sources of

pollution in the region came from the energy, industry, and

transportation sectors. Electricity generation contributed

most of the pollution in the energy sector.

The report considered four major air pollutants to be

important in the region: sulphur dioxide (SO2), nitrogen oxides

m (W.H. Leung).

.

http://crossmark.dyndns.org/dialog/?doi=10.1016/j.envsci.2013.03.010&domain=pdf

http://crossmark.dyndns.org/dialog/?doi=10.1016/j.envsci.2013.03.010&domain=pdf

http://dx.doi.org/10.1016/j.envsci.2013.03.010

mailto:[email protected]

mailto:[email protected]

http://www.sciencedirect.com/science/journal/14629011

http://dx.doi.org/10.1016/j.envsci.2013.03.010


(NOx), respirable suspended solids (RSP), and volatile organic

compounds (VOC) from hydrocarbons. In 1997, the total

emissions of SO2, NOx, RSP, and VOC in the region were

585,900, 643,000, 256,400, and 480,600 tons respectively, of

which Hong Kong produced 65,960, 123,000, 11,400, and

68,900 tons respectively. The energy sector in Hong Kong

contributed approximately 64,300, 59,700, and 6500 tons of

SO2, NOx, and RSP respectively, from generating 23,574 GWh of

electricity. The region has been one of the fastest growing

economic development areas in China, with an annual GDP

growth rate of more than 4.5% in Hong Kong (Census and

Statistics Department, 2007) and 11–12% in the PRD. Thus, the

report predicted increasing energy consumption and worsen-

ing emissions problems in the region.

In 2002, the Hong Kong government and the Guangdong

provincial government agreed to reduce, on a best-endeavor

basis, the regional emissions of sulfur dioxide, nitrogen

oxides, respirable suspended particulates, and volatile organic

compounds by 40%, 20%, 55%, and 55% respectively, by 2010,

using 1997 as the base year (HK Government news, 2002).

There was a strong demand for power companies to reduce

their emissions. Based on the agreement, the required

emission targets for Hong Kong’s electricity generation in

2010 were estimated at 25,000, 42,600, and 1260 tons of SO2,

NOx, and RSP respectively. These reduction targets were

substantial. Emissions trading was proposed as a solution for

the electricity generation industry to achieve the targets in

2002 (Liao, 2002). Emissions trading was completely new and

quite different from the emission controls that were in use at

that time. This option presented a new challenge for power

station managers who had to choose between improving their

existing equipment and/or using emissions trading to meet

the emission targets. The aim of this study was to explore a

modeling approach that would facilitate power station

managers in making such decisions.

Many different emissions trading systems have been

reported around the world (Leung et al., 2009). However, only

two major popular emissions trading programs have been

actively implemented: the Acid Rain Program based on Title IV

of the Clean Air Act adopted in the US (US EPA, 2006), and the

EU emissions trading program based on Directive 2003/87/EC

and Council Directive 96/61/EC (EC Directive, 2003) in Europe.

Both programs were developed from the ‘‘cap and trade’’

system. In HK, a pilot scheme for emissions trading among the

power stations based in HK and the PRD was also based on a

‘‘cap and trade’’ system (EPD HK, 2007a).

2. ‘‘Cap and trade’’ system

The ‘‘cap and trade’’ system is a means of preventing

emissions from exceeding government-set limits through

the trading of emission credits among the major emission

sources. An emissions monitoring station, usually the govern-

ment’s environmental protection agency, sets the air quality

objectives based on public expectations and assigns emission

caps to emissions permit holders, i.e. the power stations, who

either own or operate the emission sources. The emissions permit

holders must meet emission caps, either by improving their

equipment or by trading emission credits in the emissions

trading market. The emissions monitoring station monitors the

emissions from the emission sources, oversees the operation of

the emissions trading program, and enforces public policy by

imposing heavy penalties for any exceedance of emission

standards.

An effective ‘‘cap and trade’’ program requires legislation

so that it is enforceable by the authorities. The emission caps

must meet the air quality objectives and be respected by all of

the parties concerned. There should also be a global allocation

plan for emission caps, supported by agreement between the

emissions monitoring stations and the emissions permit holders.

The important rules in an allocation plan must be fair and

equitable and should allow for the new entry of permit

holders. The system requires a comprehensive emissions

measuring, monitoring, and reporting system. Continuous

emissions monitoring supplemented by an agreed calculation

system is commonly adopted, with heavy penalties for

exceeding limits to deter noncompliance. The pricing of

emission credits depends on the market and the cost of the

emission reduction work. Most credit trading is done through

agreed contracts that must be scrutinized by the emissions

monitoring stations.

To meet the emission caps, the emissions permit holders can

use cleaner fuels, or remove pollutants in the flue gas before

emitting it into the atmosphere. Alternatively, the permit

holders can buy emission credits to meet their emission caps,

or sell their emission credits to others when they have a

surplus. The emissions permit holders must strike a balance

between investing in emission-reduction equipment and

paying the price of emission credits. In the electricity industry,

the main concerns of the emissions permit holders are the

operating and maintenance costs, the capital investments and

the return on investment.

3. Technologies for electricity generation andpollution reduction

The emissions levels from power plants depend heavily on

which technologies the plants use for electricity generation.

The main fuels used for conventional power plants are coal

and oil. Natural gas is used in combined cycle generating

power plants.

In coal-fired or oil-fired power plants, fuels are burned in

the boiler to produce heat energy that is then converted into

mechanical energy for electricity generation. The energy

conversion efficiency for a power plant is the ratio between

the useful output of the power plant and the input in energy

terms. The theoretical energy conversion efficiency of this

process is between 35 and 42%, depending on the operating

conditions of the plant (Eastop and McConkey, 1986). With the

latest developments in technology and materials, it is possible

to achieve actual conversion efficiency of 42% at full load

(Alstom Power, 2011b).

In combined cycle plants, air is compressed in the air

compressor and gas fuel is injected into the combustion

chamber to burn with the compressed air. The combustion

process produces heat. Residual heat from the gas turbine is

used to heat water into steam in the heat-recovery steam

generator. The heat energy from the compressed air and

e n v i r o n m e n t a l s c i e n c e & p o l i c y 3 1 ( 2 0 1 3 ) 1 – 1 2 3

steam is then converted into mechanical energy for electricity

generation. The theoretical conversion efficiency is between

55 and 59% for this process (Eastop and McConkey, 1986). With

improvements in technologies and materials, it is possible to

achieve actual conversion efficiency of 59% in a combined

cycle power plant at full load (Alstom Power, 2011a). Both

processes, however, produce and emit pollutants into the

atmosphere.

3.1. Emission-reduction equipment

According to a World Bank study on clean coal technologies

(Tavoulareas and Charpentier, 1995), emissions can be

controlled by using cleaner fuel, by controlling the combustion

process, and/or by cleaning the flue gas before discharging it

into the atmosphere. As SO2 is produced when the fuel

contains sulfur, using fuel with less sulfur content produces

less SO2. SO2 emissions can also be reduced by cleaning the

flue gas after combustion with weak alkaline substances such

as limestone or lime slurry. This process can remove most of

the acidic SO2 in a flue gas desulfurization (FGD) plant.

NOx is produced when fuel burns in a combustion chamber.

Nitrogen combines with oxygen to form NOx. The amount of

NOx produced depends on the flame temperature. Low NOx

burners can reduce NOx by ‘‘staging’’ the combustion with

overfired air and reburning. Alternately, the NOx can be

cleaned using selective catalytic reduction (SCR) technology.

Ammonia is injected into the flue gas to react with NOx in the

catalytic reactor, producing molecular nitrogen and water

vapor, both of which are harmless.

Burning fuel with ash produces suspended particulates.

These particulates can be collected by an electrostatic

precipitator or by fabric filtration before discharging the flue

gas into the atmosphere. Normally, such filtration devices are

part of the boiler. An FGD plant at the exit of the boiler can also

remove the suspended particulates when the flue gas is

washed by the slurry of the reagent.

3.2. The effectiveness and cost of emission-reductionequipment

According to the World Bank study on clean coal technologies

(Tavoulareas and Charpentier, 1995), the effectiveness of

emission-reduction equipment is measured by its removal

efficiency, which is the proportion of the emissions removed

by emission-reduction equipment relative to the amount at

the inlet. The typical removal efficiencies of different

emission-reduction technologies vary according to the type

of emission. The removal efficiency of an FGD plant is about

90% for SO2, that of a low NOx burner is about 60% for NOx, and

that of electrostatic precipitation is about 99.9% for RSP

(Takahashi, 1997; US EPA, 2007). A selective catalytic reactor

can also remove about 70% of the NOx from the flue gas (US

DoE, 1997).

The capital cost of emission-control equipment depends on

the technology, design, materials, and size of the unit. The

operating and maintenance (O&M) cost of such equipment is

dominated by the cost of the materials and labor required to

operate the machines. This cost also depends on the amount

of emission reduction that the equipment can achieve.

Tavoulareas and Charpentier (1995) reported that the capital

and O&M costs for an FGD plant were US$ 100–150/kW of

capacity and US¢ 0.15–0.33/kWh of electricity generated,

respectively. The capital and O&M costs for a low NOx burner

were US$ 20–50/kW of capacity and US¢ 0.5–1/kWh of

electricity generated, respectively, while those for an SCR

plant were US$ 45/kW of capacity and US$ 2165/tons of NOx

respectively. In the PRD, the equipment price for retrofitting

FGD to a group of power plants was about US$ 50/kW of

capacity, derived from the ‘‘Shandong flue gas desulfuriza-

tion’’ project in China (World Bank, 2007).

4. Air pollution control and emissions tradingframework in the region

In HK, emission control is enforced under the Air Pollution

Control Ordinance (Chapter 311 of the laws of the Hong Kong

Special Administrative Region) (Ordinance HK, 1997) and the

Air Pollution Control (Specified Processes) Regulation (Regula-

tion HK, 2007). The major stationary air polluters, such as

power plants, are subject to stringent emission controls. Their

licenses specify all of the emission limits for their operations,

which are normally given in terms of allowable concentrations

of emissions. Emission caps were imposed on power plants in

2006.

In 1997, the Hong Kong government ceased to allow the

construction of new coal-fired generating units in favor of

natural gas plants (EPD HK, 2011). Only renewable energy or

natural gas-fired combined cycle plants were considered

thereafter. The renewable energy options for electricity

generation include hydropower, solar, wind, geothermal,

biomass, nuclear, or ethanol (US Energy, 2005). The emissions

from these types of plants are zero (EMSD, 2011).

In the PRD, several laws have been enacted to control

emissions. The principal law is the ‘‘Law of the People’s

Republic of China on the prevention and control of atmo-

spheric pollution’’ (Law of PRC, 2000). According to this law,

the central government sets nationwide standards for emis-

sions. Local governments can set their own requirements for

local standards of emission control. The central authorities

delegate power to local governments for setting any standards

not covered by the national regulations, or for making local

standards more stringent than their national counterparts. In

addition to these emission-control laws, there are national

standards to govern emissions such as the ‘‘GB13223-1996

emission standard of air pollutants for thermal power plants’’

(GB China, 1996). Local governments have also set up various

regulations, such as the city of Shenzhen’s ‘‘Regulations on

controlling acid rain and emissions of sulfur dioxide in

Shenzhen’’ (Regulation SZ, 1998).

An emissions trading agreement, entitled the ‘‘Emissions

trading pilot scheme for thermal power plants in the Pearl

River Delta region,’’ has been signed by the Hong Kong and

Guangdong provincial governments (EPD HK, 2007a). Since

then, emissions trading scheme has been executed in HK and

PRD. However, both the cleaner fuel and the retrofitting emission-

reduction equipment approaches can meet emission targets

from 2010 to 2016 in Hong Kong. Emissions trading is only

considered as a market-based fall-back option, in case of an

Demand forecast

Genera�on mach ines

Coal -firedNatural gas -fired

Emiss ions

Emiss ion caps

Compare with cap

Sell

Emiss ions trading market

Improve the exis�ng equ ipment

Buy

Produ ce electricity

Cost of bu ying emiss ion all owanc e

Cost of capital and O&M

Fig. 1 – The details of the emissions trading model (ETM).


unexpected emergency in normal operations. This scheme is

intended to support the trading of SO2 emission credits and

also includes the trading of NOx and RSP emissions. This is a

‘‘cap and trade’’ system on a project basis. An emissions permit

holder who is interested and eligible may propose emission

reduction plans that would then be deemed acceptable by the

environmental protection authorities. The surplus between

the base emissions targets before and after the completion of

an emission reduction plan can be sold as ‘‘project-based

emission credits.’’ All participating power stations must

install suitable systems to monitor their emissions. Emission

credit trading is a market activity in the form of a business

contract that must be submitted to the environmental

protection authorities for approval. Once the authorities verify

the actual total emissions for a year, any emission credits can

then be transferred. If the seller fails to transfer the emission

credits, then he/she must compensate the buyer according to

the contract. To enforce overall compliance with regional

pollution limits, the central and local governments prosecute

violators of emission caps in accordance with their laws and

regulations.

5. Business environment of Hong Kong’selectricity generation industry

Electricity generation in Hong Kong is controlled by a business

model referred to as the Scheme of Control Agreement, which

applies to emissions permit holders. The government controls

the investments and profits of the emissions permit holders in

accordance with the Scheme, which is designed to ensure a

safe and reliable supply of electricity at an affordable price. A

new agreement was signed on January 1, 2009 to cover the next

10 years, with a possible extension for a further five years (EB

HK, 2008a, 2008b). Linear depreciation of fixed assets is

adopted. The permitted rate of return on electricity generation

is 9.99% on average net fixed assets and 11% on average

renewable net fixed assets. Incentives or penalties on the

permitted return are also given, based on environmental

performance, customer performance, energy efficiency, and

renewable energy production. Hence, HK’s electricity market

is still regulated, and will remain so for at least several years.

6. Emissions trading model (ETM) forelectricity generation with multi-fuel operationsin Hong Kong

To analyze the complicated emissions and electricity genera-

tion industry in Hong Kong, the characteristics of HK’s

electricity industry and the emissions trading program are

integrated into the Emissions Trading Model (ETM) for Hong

Kong, which applies mainly to natural gas-fired and coal-fired

power plants.

The driving force of the industry is the demand for

electricity. Power plants try to meet the demand for electricity

while controlling the emissions they produce in the process.

The emissions are reduced either by using clean coal

technologies or by trading emission credits to meet the

emission caps. The ETM is designed to capture the factors

involved, as Fig. 1 highlights.

6.1. The structure of the ETM

The ETM is separated into four modules: (1) an equipment

module, (2) a load dispatching module, (3) an emission

module, and (4) a financial module.

6.1.1. The equipment moduleThe equipment module is primarily based on the technology

of electricity generation. Two major types of electricity

generation equipment are considered: the combined cycle

gas turbine plant burning natural gas and the coal-fired plant.

The modules created for combined cycle gas turbine and coal-

fired power plants are shown in Fig. 2.

The emission performance of a power plant is measured by

its emission factor. When the emission performance is better,

the emission factor is lower. The emission factor for a specific

emission is measured by the amount of emission produced

divided by the amount of electricity generated (US EPA, 2011).

The emission factors for SO2, NOx and RSP for coal-fired and

combined cycle power plants can be calculated using this

equipment module.

Combined cycle power plants use natural gas, which

contains very little sulfur and ash, hence SO2 and RSP

emissions are close to zero. NOx emissions can also be

controlled by using a dry low NOx burner in the combustion

chamber. The energy conversion efficiency of a combined

cycle power plant is high, at around 55–59%. The emission

factors are low and the emission performance of the power

plant is high (GE Energy, 2011).

Coal-fired power plants use coal, which contains higher

sulfur and ash, thus SO2 and RSP are produced after

combustion. The combustion process also produces NOx.

The energy conversion efficiency of this type of power plant is

lower than that of a combined cycle plant, at around 35–42%.

The emission factors are higher and the emission perfor-

mance is lower than for a combined cycle plant.

Fig. 2 – The combined cycle gas turbine and coal-fired power plant modules.


6.1.2. The load dispatching moduleThe demand for electricity is allocated to different power

plants according to priority, which can be determined either

by the efficiency of the power plant or by its emissions

performance. Under stringent emission control, the electricity

demand will be logically dispatched according to emission

performance: a gas-fired combined cycle power plant will have

priority over a coal-fired power plant. For power plants with

the same emission performance, a plant with higher efficiency

will have priority over a less efficient one.

Each machine bears a load factor, which is the ratio of the

actual annual electricity generated to the total annual possible

production capacity. Hence, the actual electricity production

can be estimated by multiplying the total annual possible

capacity by the load factor.

The total emissions can be predicted by multiplying the

actual electricity generated by the emission factor, which can

Fig. 3 – The load disp

be obtained from the ‘‘equipment module.’’ Each emission for

a power plant is then summed to find the total emission of the

power station. In this module, the total emission of each

pollutant is estimated according to the electricity generated,

the emission factors of the machines and the mode of load

dispatching.

The module created for load dispatching is shown in Fig. 3.

6.1.3. The emissions module

In the emissions module, the total emission of each pollutant

is compared with the emission cap. If the emission caps are

not exceeded, then nothing needs to be done and the surplus

emission credit can also be sold in the emissions trading

market. If the total emissions exceed the emission caps, then

the emissions permit holder can either buy emission credits from

the emissions market, or try to improve its emissions

performance. The emissions module is shown in Fig. 4.

atching module.

Emission cap s

SO2, NOx, RSPIf actual

> cap

Do no thing or

sell emiss ion s

allowance

yes

Improve the

existing

equipmen t

Trade

emiss ion

credits

no

The emiss ion s modu le

Emissions

SO2, NOx, RSP

Emiss ions

SO2, NOx, RSP

Emiss ions

SO2, NOx, RSP

Sum of each emiss ion , SO2, NOx, RSP

Fig. 4 – The emissions module.


6.1.4. The financial moduleIf the actual emissions exceed the emission caps, the emissions

permit holder must either improve its equipment or buy

emission credits. Both options involve financial costs. Im-

proving the equipment includes upgrading the existing

installation or adding new emission-reduction equipment.

This option requires capital investment and the future O&M

cost of the equipment, hence the financial module of the ETM

is separated into two parts: (1) improving the existing plant by

installing emission-reduction equipment and (2) trading

emission credits with other emissions permit holders.

6.1.4.1. Financial module for improving existing equipment. In-

stalling emission-reduction equipment requires capital

Fig. 5 – The finan

investment. This capital cost can be obtained by multiplying

the capacity of the power plant by the capital cost of the

equipment per unit capacity as given by Tavoulareas and

Charpentier (1995). This capital investment generates two

capital expenses, the depreciation of the equipment and the

O&M costs incur to operate and maintain the new equipment.

A linear depreciation of capital cost is adopted. The O&M costs

are calculated by multiplying the amount of electricity

produced by the O&M costs per unit electricity produced as

given by Tavoulareas and Charpentier. The operating cost for

emission-reduction is the sum of the depreciation cost and the

O&M costs of the additional emission-reduction equipment.

The predicted price of an emission is calculated from these

costs which can be used as the price to trade the emission

credits with the other power stations. Since different power

plant has different ‘‘predicted price of the emission credit’’,

this difference creates the incentive to trade among the power

stations. In determining whether to ‘‘trade’’ or not in the

emissions trading market, the emissions permit holder would

compare the predicted price of the emission credit to the price

of buying the emission credit in the market for a period of

time.

6.1.4.2. Financial module for trading emission credits. The only

operating cost for this module is the cost of buying the

emission credits. If there is an emission trading market, the

price of emission credit can be determined by the trading

market. Since the actual emissions trading market was not

exercised, the price of the emission credit had to be

estimated. The price of emission credit is based on the

principle of the pricing in the emissions market, i.e. the

difference average operating cost of an emission reduction

equipment between HK and PRD according to the cap and

trade system (US EPA, 2006). In Hong Kong, the international

price of the emission reduction equipment is used, which is

derived from the research data published by the World Bank

(Tavoulareas and Charpentier, 1995). However, in PRD the

cial module.

z = total cost of emission

reduction for a power station

i = an emission

F = cost of buying emission

credits

I = all emissions

C = the cost of operating the

emission-reduction equipment

m = price of emission credit

E = electricity generated

e = quantity of emissions

a = additional emission-

reduction equipment

i (cap) = emission cap

T = the total years of study

after the base year

a = operating and

maintenance cost per

unit of emission reduction

1 = the base year

b = rate of depreciation

of the capital cost

t = a year after the

base year

C = capital cost of emission-

reduction equipment

r = discount rate

j = removal efficiency

NPV = net present value

s = emission per unit of

electricity generated

(emission factor)

n = power plant of the

power station

F = fuel specification and

N = all power plants

of the power station

h = conversion efficiency

of the power plant


home market price is used, which is derived from a project on

emission reduction submitted by the Chinese Government to

the World Bank (2007). The price in HK is higher than that in

PRD. Hence, it is possible to trade the emissions between HK

and PRD.

For a power station, the total operating cost of the emission

reduction is the sum of the cost of operating the additional

emission equipment and the cost of buying the emission

credits if the limit is exceeded.

The financial module is shown in Fig. 5.

6.2. Linear programming method for optimization

The ETM model applies to the multi-variables to meet multi-

targets, i.e. the emission caps (e(cap)) on the emissions (i) for

electricity generation. The linear programming (LP) method

can be used to describe the cost/benefit optimization process.

The objective function of the LP model is the total operating

cost of the emission reduction z(ET, aT) for (N) power plants

with two variables of electricity generated (Et) and the

additional emission-reduction equipment for each power

plant (at) in a period (T). In this model, Et in each period t is

given based on electricity demand forecast. The total operat-

ing cost z(Et, at) is composed of the net present value (NPV) at a

discount rate (r) for two elements: the cost of buying emission

credits (F) and the cost of operating the emission-reduction

equipment (C) to meet the emission caps (e(cap)) from the base

year (1) to year (T). Both F and C are functions of the electricity

generated (Et) and the additional emission-reduction equip-

ment (at).

The cost of buying emission credits F (Et, at) is the sum of

the price of emission credits m (ai,t) times the emission

exceedance ([P

ei (En,t, an,i,t) � ei (cap),t]) for each emission. The

cost of operating the emission-reduction equipment C (Et, at) is

the sum of O&M cost which is calculated by multiplying

electricity produced (Et) by the O&M cost per unit electricity a

(an,i,t) plus the depreciation cost which is calculated by

multiplying the capital cost C (an,i,t) with the rate of deprecia-

tion b (an,i,t).

To determine whether to purchase emission credits or not,

it is necessary to minimize the operating cost of emission

reduction for a period (T). However, it has to meet the

constraint that there is no exceedance of emission caps with

the period (T) for all emissions (I). The following is the LP model

function for the ETM.

To minimize

zðET; aTÞ ¼XT

t¼1

NPVfrt; ½FðEt; atÞ þ CðEt; atÞ�g

with function F (Et, at) is equal to

with FðEt; atÞ ¼XI

i¼1

mðai;tÞ �XN

n¼1

eiðEn;t; an;i;tÞ � eiðca pÞ;t

" #

and CðEt; atÞ ¼XI

i¼1

XN

n¼1

f½aðan;i;tÞ � En;t� þ ½bðan;i;tÞ � Cðan;i;tÞ�g

while eiðEn;t; an;i;tÞ ¼ ½1 � jðan;i;tÞ� � sðFn; hnÞ � En;t

provided that ½eiðca pÞ;t � eiðEt; ai;tÞ� 3 0 for

ð1 � t � TÞ \ ð1 � i � IÞ

where

6.3. Different business strategies for the emissionstrading program

The electricity generation in the study period is based on the

predicted growth rate of the electricity demand starting from

the base year. For the provision of additional emission-

reduction equipment, there are many possible combina-

tions, i.e. a multiple of the number of power plants and

emissions. To simplify the simulation, four possible business

strategies are considered to meet the new emissions

requirements.

Option 1: Emissions trading strategy: rely on the emissions

trading market to buy and sell emission credits to

meet the emission caps. It is not necessary to install

any emission-reduction equipment.

Option 2: Mixing strategy: install limited emission-reduction

equipment for SO2 and NOx and buy emission

credits to cover any exceedance of emission limits.

This is a compromise between Option 1 and

Option 3.

Option 3: Emission reduction strategy: install just enough emis-

sion-reduction equipment to meet the emission

caps with normal outage for maintenance.

Option 4: Secured emission reduction strategy: install enough

emission-reduction equipment to meet the emis-

sion caps with allowance for a single serious

equipment failure per year.

6.4. Simulation of the ETM

The four modules of the ETM are simulated in a computer

program to calculate the total operating cost, emissions, and

the capital investment for each option. The boundary

conditions of the simulation are as follows:


(1) The period of simulation: 2006–2016.

(2) The model simulates SO2, NOx, and RSP emissions because

these are the pollutants of concern in the region.

(3) The characteristics of the power plants as an emission

permit holder.

The input variables are (1) the electricity demand forecast

from 2006 to 2016 and (2) the additional emission-reduction

equipment to be installed in a specific year for each option.

The outputs from the computer program are (1) the total

emissions for each year, (2) the capital cost required for

additional emission-reduction equipment, and (3) the total

operating costs of the new equipment.

7. Simulation of the ETM, 2006–2016

The power plants of an emissions permit holder in Hong

Kong, Company A, are chosen to demonstrate the simula-

tion. Company A has a total installed capacity of 6908 MW.

The company operates four 350 MW and four 677 MW

conventional coal-fired units that have been fitted with

low NOx burners and electrostatic precipitators, and

eight 312.5 MW gas-fired combined cycle units. The total

electricity generated in 2006 was 26,400 GWh (CLP, 2007;

PAH, 2011).

The simulation starts in 2006, 10 years after the base year of

1997 when the new emission controls were introduced in HK.

The following parameters are input to the program: (1) the

characteristics of electricity generation machines including its

capacity, type, emission-reduction facilities if any, (2) the

depreciation requirement from the ‘‘scheme of control’’

agreement, (3) the emission caps based on the emission

reduction agreement of 2002 between HK and the PRD, and (4)

the emissions trading pilot scheme for thermal power plants

in the HK-PRD region.

Based on the historical data, the growth in electricity

generation was about 1.5% from 2000 to 2006 (CLP, 2007; PAH,

2011). This growth rate is used to project the electricity

demand from 2006 to 2016. The authority allocated the

emission caps for Company A up to 2009 (EPD, 2005, 2007b),

after which the emission caps were estimated according to the

agreement on air quality (CH2M HILL, 2002).

The electricity demand forecast and the predicted emission

caps for Company A are shown in Table 1.

The international prices were used for (1) the equipment

capital cost relating to the capacity of the unit and (2) the

operation and maintenance (O&M) costs per unit of emission

reduction for retrofitting emission-reduction equipment. The

Table 1 – The estimated electricity generation and the predict

Year 2006 2007 2008 2009

Electricity production (GWh) 26,408 26,804 27,204 27,608

SO2 emission cap (T/yr) 50,000 46,000 41,400 39,400

NOx emission cap (T/yr) 30,400 28,800 27,660 27,400

RSP emission cap (T/yr) 2030 1710 1165 1115

inflation rate and the discount rate in HK was about 3% for the

period (Trade Economics, 2012).

7.1. The simulation of the LP model for an emissionspermit holder from 2006 to 2016

Based on the above estimations, a simulation of the electricity

generation for Company A from 2006 to 2016 was conducted.

The results of the simulation are as follows.

For Option 1, emission allowances for SO2, NOx, and RSP are

purchased in the emissions market to meet the emission caps.

No other capital or operating costs are incurred. The

simulation is applied to SO2, NOx, and RSP emission reduction

separately.

The results of the simulation show that there is no

exceedance of the SO2, NOx, or RSP caps from 2006 to 2008

because the emissions are below the levels allowed by the

emission caps. However, beginning in 2009, the SO2 and RSP

emissions exceed their respective emission caps, and NOx

emissions exceed the emission cap from 2010 onwards. Hence,

Company A has to buy emission credits. The results of this

simulation are shown in Table 2.

For Option 2, only limited emission-reduction equipment

is provided to reduce emissions, with some allowance for

emission exceedance. Company A provides one FGD system

for one 677 MW coal-fired power plant per year in 2009 and

2010, and one selective catalytic reactor for one 677 MW

coal-fired power plant in 2010. The emission exceedance is

covered by buying credits from the emissions trading

market.

The results of the simulation for Option 2 are summarized

in Table 3.

In Option 3, no exceedance of emissions is allowed.

Company A needs to install one FGD system in one of its

677 MW coal-fired units in 2009, two FGD systems in 2010, and

one in 2011, with one selective catalytic reactor for each of two

677 MW coal-fired units, the first in 2010 and the second in

2011.

The results for the simulation for Option 3 are summarized

in Table 4.

In Option 4, the company has to provide one more

emission-reduction system in addition to those installed in

Option 3. Therefore, the company needs to install two FGD

systems in two of its 677 MW coal-fired units in 2009 and 2010,

respectively, one FGD in one of its 350 MW coal-fired units in

2011, and two selective catalytic reactors in two 677 MW coal-

fired units in 2010.

The simulation results for Option 4 are summarized in

Table 5.

ed emission caps for Company A (2006–2016).

2010 2011 2012 2013 2014 2015 2016

28,016 28,428 28,844 29,264 29,688 30,115 30,547

14,850 14,850 14,850 14,850 14,850 14,850 14,850

26,400 26,400 26,400 26,400 26,400 26,400 26,400

745 745 745 745 745 745 745

Table 3 – The simulation results on emission exceedance costs for Company A, 2006–2016 (Option 2).

Emission Items Unit Year Total

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

SO2 Produced Tons 36,591 40,830 41,070 32,741 23,416 26,642 29,436 31,008 32,538 34,139 35,754 363,857

Exceedance Tons 0 0 0 0 8531 11,757 14,551 16,123 17,654 19,254 20,869 108,430

Capital Cost HK$ million 0 0 0 316.8 316.8 0 0 0 0 0 0 633.7

Total operating cost HK$ million 0.0 0.0 0.0 62.1 117.7 133.0 145.7 154.4 163.0 172.8 181.8 1448.7

NOx Produced Tons 23,897 24,499 25,104 25,716 22,541 23,167 23,799 24,434 25,052 25,700 26,354 270,263

Exceedance Tons 0 0 0 0 0 0 0 0 0 0 0 0

Capital Cost HK$ million 0.0 0.0 0.0 0.0 237.6 0.0 0.0 0.0 0.0 0.0 0.0 237.6


RSP Produced Tons 1492 1391 1104 909 715 750 786 821 856 892 929 10,745

Exceedance Tons 0 0 0 0 0 8 43 78 113 149 186 577


Table 2 – The simulation results of emission exceedance costs for Company A, 2006–2016 (Option 1).


2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

SO2 Produced Tons 36,591 40,830 41,070 43,425 44,741 49,521 53,331 54,855 56,338 57,891 59,459 538,051

Exceedance Tons 0 0 0 4025 29,856 34,636 38,446 39,970 41,453 43,006 44,574 275,390

Cost of exceedance HK$ million 0.0 0.0 0.0 23.6 173.6 185.7 195.3 201.4 207.4 215.2 224.0 1426.3


Exceedance Tons 0 0 0 0 0 556 1189 1824 2442 3090 3744 12,840


RSP Produced Tons 1492 1391 1104 1138 1173 1208 1243 1278 1313 1349 1386 14,074

Exceedance Tons 0 0 0 23 430 465 500 536 570 607 643 3774


e n

v i

r o

n m

e n

t a

l s

c i

e n

c e

&

p

o l

i c

y 3

1 (

2 0

1 3

) 1

– 1

2

9



2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

SO2 Produced Tons 36,591 40,830 41,070 32,741 23,416 26,642 29,436 31,008 32,538 34,139 35,754 363,857

Capital Cost HK$ million 0 0 0 316.8 633.7 316.8 0 0 0 0 0 1267.3





RSP Produced Tons 1492 1391 1104 909 487 366 376 387 399 435 471 7818




2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

SO2 Produced Tons 36,591 40,830 41,070 22,057 6623 7357 8005 8322 8637 8967 9305 197,764






RSP Produced Tons 1492 1391 1104 681 355 365 376 386 397 407 418 7373


e n

v i

r o

n m

e n

t a

l s

c i

e n

c e

&

p

o l

i c

y 3

1 (

2 0

1 3

) 1

– 1

21

0

Table 6 – The total emissions, capital cost and operating costs for Company A (2006–2016) for all four options.

Item (unit) SO2 produced(tons)

NOx produced(tons)

RSP produced(tons)

Capital expenditure(HK$ million)

Total operatingcost (HK$ million)

Option 1 538,051 296,811 14,074 0.0 1695.2

Option 2 364,166 270,263 10,645 871.3 1441.8

Option 3 218,987 270,263 7818 1505.0 1366.9

Option 4 197,764 243,715 7373 1906.4 1876.4

0.0

500.0

1,000 .0

1,500 .0

2,000 .0

2,500 .0

0

100,000

200,000

300,000

400,000

500,000

600,000

Op�on 1 Op �on 2 Op �on 3 Op �on 4

Expe

nses

(HK$

mill

ion)

Emis

sion

s (to

ns)

Total Emissions and E xpenses fo r 4 Op�ons for 10 Ye ars

SO2 NOx RSP Capital Expend iture

Fig. 6 – The total emissions, capital cost, and operating costs for Company A, 2006–2016, for all four options.


7.2. Summary of results for the four options from 2006 to2016

Based on the simulation results for the four options, the total

operating cost of the equipment, the total capital investment,

and the total emissions of SO2, NOx, and RSP from 2006 to 2016

are summarized in Table 6.

These results are plotted in Fig. 6.

8. Discussion and conclusion

According to the objective function graph in Fig. 6, the total

operating cost decreases from Option 1 to Option 3 and

increases in Option 4. The operating cost of Option 3 is the

lowest among the four. Option 3 produces a reduction of about

60% in SO2 and 9% in NOx compared with Option 1. The capital

costs increase from Option 1 to Option 4 as more emission-

reduction equipment is installed. In Option 4, both the capital

and operating costs are higher than those of Option 3, but the

reduction in emissions is not significant. Hence, Option 3 is the

most reasonable choice.

In early 2005, Company A proposed (1) to commission new

combined cycle power plants in 2005 and 2006 and (2) to

retrofit the four 677 MW coal-fired units with FGD and SCR

plants to reduce atmospheric emissions (EDLB, 2005). The

company announced that it had achieved its environment-

related goals. The company recorded a material reduction of

60% for SO2, NOx, and RSP. Therefore, the company’s emission

control efforts are consistent with the recommendation of

this study.

With the market forces and the flexibility of emission

trading, power plants can use this emission trading model

proactively to work out different cost-effective emission

reduction strategies to benefit the whole region. In reviewing

the trade or no trade decision, power stations could use the

model as reference guides to monitor the effectiveness of their

emission reduction strategies so that they could make proper

and timely investment decisions in reducing emission. The

emission trading model can be used to simulate various

business alternatives for power stations with multi-fuel

operations to meet the stringent emission controls at the

lowest operating costs. The research can also be extended to

analyze a complicated multivariable and multi-constraint

scenario for different large electricity companies with multi

power stations and plants. This in term would facilitate the

implementation of an effective emissions trading scheme in a

region.

Leung et al. (2009) commented that simplicity, flexibility,

periodic review, and broad participation are the important

factors for the successful implementation of an emission

trading scheme. The government policy makers can use this

emission trading model to evaluate the effectiveness of the

emission trading scheme in a region. With the simplicity of the

model, many different scenarios, including multi-power

stations and plants in a region, can be simulated to provide

useful information for policy makers to assess the practicality

of the emission trading scheme. This would help them to

formulate realistic emission trading strategies for the whole

region.

While the emission trading simulation model can provide

results that are difficult to be experimentally measurable, yet,


in applying it to the complicated emission trading scenario,

the model is required to be interpreted with cautions. The

simulation of the emission trading model involves the

interaction of a number of variables; the relationship of these

variables could be quite complicated. In a steady electricity

growth situation, the linear functions could be approximated.

However, under a highly fluctuated environment, the adop-

tion of the model may require special care. To enhance the

model, future research could try to investigate situations

where non-linear functions may apply.

r e f e r e n c e s

Alstom Power, 2011a. Key reference of Belchatow pulverizedcoal steam unit of 858 MW, website of Alstom Power.

Alstom Power, 2011b. KA26 combined cycle power plant,website of Alstom Power.

Census and Statistics Department, HK, 2007. Statistical data onGDP from census and statistics department of HKSAR,website of the Census and Statistics Department of HKSAR.

CH2M HILL (China) Limited, 2002. Final report, Agreement No.CE 06/98. Study of air quality in the pearl river delta region,CH2M Hill (China) Limited in association with ENSRInternational, Inc.; Rowan William Davies and Irwin, Inc.;Guangdong Province Environmental Monitoring Centre;Guangzhou Research Institute of Environmental Protection.

CLP Power (CLP), 2007. CLP group – about us, website of CLP.Eastop, T.D., McConkey, A., 1986. Applied Thermodynamics for

Engineering Technologists. Longman Group, UK, 232–242.EC Directive, 2003. Directive 2003/87/EC of the European

Parliament and of the Council of 13th October 2003establishing a scheme for greenhouse gas emissionallowance trading within the Community and amendingCouncil Directive 96/61/EC, Official Journal of the EuropeanUnion, 25th October 2003 Edition.

Economic Development and Labour Bureau (EDLB), HKSAR,2005. The CLP Power Hong Kong Limited Castle Peak PowerCompany Limited 2005 Financial Plan, the LegislativeCouncil Panel on Economic Services, Legislative Council,Hong Kong.

Electrical and Mechanical Services Department (EMSD) ofHKSAR, 2011. HK RE Net – overview, website of HK RE Net,EMSD, HKSAR.

Environment Bureau (EB HK), HKSAR, 2008a. The Scheme ofControl Agreement between HKSAR and CLP, 2008, websiteof the Environment Bureau, HKSAR.

Environment Bureau (EB HK), HKSAR, 2008b. The Scheme ofControl Agreement between HKSAR and HEC, 2008, websiteof the Environment Bureau, HKSAR.

Environmental Protection Department (EPD HK), HKSAR, 2005.Press Release – Power Station’s Licence Renewed withEmission Caps, website of Environment ProtectionDepartment, HKSAR.

Environmental Protection Department (EPD) of HKSAR, 2007a.Implementation Framework of the Emission Trading PilotScheme for Thermal Power Plants, website of EnvironmentalProtection Department, HKSAR, website of EnvironmentProtection Department, HKSAR.

Environmental Protection Department (EPD) of HKSAR, 2007b.Press Release – Power Station’s Licence Renewed withTightened Emission Caps, website of EnvironmentProtection Department, HKSAR.

Environmental Protection Department (EPD) of HKSAR, 2011.Climate Change – Electricity Generation by Fossil Fuel, thewebsite of Environmental Protection Department, HKSAR.

GB China (GB), 1996. GB13223-1996, Emission Standard on AirPollutants for Thermal Power Plants. National Standard ofthe People’s Republic of China, GB.

General Electric Energy (GE Energy), 2011. General ElectricEnergy on Advanced Technology of Combined Cycle, thewebsite of General Electric Energy.

HK Government news (Hong Kong SAR press release), 2002. JointStatement on Improving Air Quality in the Pearl River DeltaRegion, website of the HKSAR.

Hong Kong Observatory (HKO), 2007. Meteorological Statistics –Number of hours of Reduced Visibility (Hong KongObservatory), website of the Hong Kong Observatory.

Law of PRC, 2000. Law of the People’s Republic of China on thePrevention and Control of Atmospheric Pollution, 2000, Lawsof the People’s Republic of China; adopted 1995; effective 2000.

Leung, D.Y.C., Daniel, Y., Amanda, N., Leung, M.K.H., Alan, C.,2009. An overview of emissions trading and its prospects inHong Kong. Environmental Science & Policy 12 (February (1))92–101.

Liao, S., 2002. Speech in Legislative Council on EmissionsTrading. In: Minutes of Legislative Council Meeting on 6thNovember 2002, Legislative Council, Hong Kong.

Ordinance HK (Laws of Hong Kong), 1997. Air Pollution Control(Specified Processes) Regulations, Hong Kong Ordinance,Version date: 30 June 1997.

Regulation HK (Laws of Hong Kong), 2007. Chapter 311, AirPollution Control Ordinance, website of the Department ofJustice, HKSAR.

Takahashi, M., 1997 August. Technologies for reducingemissions in coal-fired power plants. In: Energy Issue 3,World Bank.

Regulation SZ (Environmental Protection Department ofShenzhen), 1998. Regulation on Controlling Acid Rain andEmission of sulfur Dioxide in Shenzhen.

Tavoulareas, E.S., Charpentier, J.-P., 1995. World Bank, CleanCoal Technologies for Developing Countries. In: The WorldBank Technical Paper Number 286.

The Power Assets Holdings Limited (PAH), 2011. Power AssetsHolding, website of Power Assets.

Trade Economics, 2012. China Inflation Rate, website of TradeEconomics.

US Department of Energy and Southern Company Services, Inc.,1997. Control of Nitrogen Oxide Emission: Selective CatalyticReduction (SCR), website of the US Department of Energy, USGovernment.

US Energy (US Energy Information Administration), 2005.Official Energy Statistics from the U.S. Government, websiteof Official Energy Statistics from the US Government.

US EPA, 2006. Acid Rain Program SO2 Allowance Fact Sheet,website of the Office of Air and Radiation Clean Air MarketDivision.

US EPA, 2007. Electrostatic Precipitators, website of the US EPA,US Government.

US EPA, 2011. Emission Factors & AP42, website of the US EPA,US Government.

World Bank, 2007. China – Shandong Flue Gas DesulfurizationProject (2007), website of the World Bank.

Richard C.M. Yam is an associate professor in the Department ofSystem Engineering and Engineering Management at City Univer-sity of Hong Kong. His current research interests are in the areas ofproduct innovation and technology management.

W.H. Leung is a professional engineer who has worked in a HongKong power station for over 30 years. He is also an engineeringdoctoral student at the City University of Hong Kong. His currentresearch interests are in the areas of environmental protectionand engineering management.

http://refhub.elsevier.com/S1462-9011(13)00068-3/sbref0030














emissions trading in hong kong and the pearl river delta region

Documents