Chapter NCC 1027

1

Assessment of the policy of “promoting large and closing small” in

China’s power sector in 2006-2013：a perspective of CO2 mitigation

based on vintage structure

Shuwei ZHANG*,1, Xuying Qin2

1Draworld Environment Research Center (DERC), 100101, Beijing, China; 2Nuclear and New 5

Technology Institute (INET), Tsinghua University, 100084, Beijing, China

Abstract

“Promoting large and closing small”, i.e. decommissioning the old, small and inefficient

power plants with large scale units (larger than 600 MW) is one policy in China’s power

sector adopted in 2006-2013, to promote the energy efficiency and emission mitigation. The 10

scale of nearly 90 GW early-retirement and massive new capacity installation in the same

period alter the aggregated pattern of coal power fleet significantly. In this paper, we

measure the effects of this large scale closure on the vintage structure, efficiency and

cumulative emission of CO2 with stock nature, upon building the vintage structure of coal

power installations methodologically based on stock variables and assumptions on the 15

distribution of retirement units. Considering coal limitation policy by 2020 or not, the

environmental outcome of this policy is different in attribute due to long duration of power

plant and the depreciation schedule. 1.2 billion more emission during 2005-2050 occurred

if coal is banned from 2020, and if no, 0.1 billion reduction can be obtained as a result of this

policy. Finally, the merit of this policy was assessed based on common environmental 20

economics standards, and the likely implementation of banning coal by 2020 was discussed.

*Corresponding author. Tel.: +86 10 83607299; fax: +86 10 84872259. E-mail address: [email protected].

Present address: Floor 12, Block B, Locker Time Center, No.103 Huizhongli, Chaoyang District, 100101,

Beijing

2

Long-term consequences of current energy investment and policy choices should be

considered explicitly in the context of plausible climate mitigation policy.

Keywords: China; Power Sector; Early retirement; Vintage capacity; CO2 emission

3

Contents

1. Introduction .................................................................................................................. 4

2. Vintage capacity structure and calibration ........................................................ 5 5

2.1 Equations and data flow ...................................................................................................... 5

2.2 Historic stock and its efficiency ........................................................................................ 7

2.2 Assumptions on the vintage-distribution of early-retirement (ER) units ....... 9

2.3 Efficiency of new additions ............................................................................................. 10

3. Scenarios...................................................................................................................... 11 10

4. Results .......................................................................................................................... 12

4.1 Different patterns of coal power fleet in 2013 ........................................................ 12

4.2 Disparity of total coal capacity ....................................................................................... 13

4.3 Cumulative emission difference .................................................................................... 14

4.3 Good or bad policy? ............................................................................................................ 15 15

5. Discussions ................................................................................................................. 18

5.1 CCS not game changer ....................................................................................................... 18

5.2 Envision of the possibility of coal limitation from 2020 ...................................... 18

6. Conclusion ................................................................................................................... 19

References ........................................................................................................................... 20 20

Table ...................................................................................................................................... 21

Figures ................................................................................................................................... 21

4

1. Introduction

Since 2005, China said wave to the long lasting power supply tight situation, and have

started to implement the policy of using high efficiency, large scale coal-thermal units

(larger than 600 MW) to replace the old and inefficient ones, on the context of seeking

better energy efficiency in power sector and achieving the stringent 20% energy intensity 5

decline target (Zhang and Bauer, 2013). This policy, so-called “promoting large and closing

small”, resulted in a shutdown of small units about 77 GW in the 11th Five-year-plan (FYP)

from 2006-2010, and further about 10 GW from 2011-2013 (NEA, 2014). This policy lifted

China’s power sector to large units dominating era in power mix. Up to now, the 600 MW

and above size units, is over 1/3 in the thermal power mix. As a comparison, in the year of 10

2005-2006, units with size smaller than 200 MW are as much as 47% in the power fleet

(Ouyang, 2014).

Much research has been done for the accounting of the emission reduction implication of

China’s phase-out of small power plants and energy equipments in the past 11th FYP (e.g.

CEC, 2011; Price et al., 2011; Wang and Chen, 2010). They found that significant energy 15

saving and emission (whatever local and global polluters) reduction achieved,

benchmarking to a constant efficiency level in 2005.

Contrast to the static measurement, considering the cumulative emission might be another

picture. Compared to the small power plants built at 1990s, the new-built large plants will

usually run another 3-4 decades due to sunk capital cost, which means substantially more 20

cumulative CO2 emission (“committed emissions” defined at Davis and Socolow, 2014 and

Davis, 2010) than the small plant in the remaining years (“remaining commitment”). When

lifespan of old plants past in some future point, new coal-fired plant construction might be

not applicable with policy restriction and/or competition from other options. It is a solution

to be equipped with carbon capturing and storage (CCS) to reduce the emission along the 25

life time. But from current perspective, this is an unproved technology with inherent risk,

and costly especially when retrofitting the existing plants. This also implies the possibility of

large-scale stranded coal power assets because of uncertainty in climate policy.

Rooted in this story, the research question in this paper is “what are the long term impacts

of the early-retirement of small power units in china, compared to the counter-factual case 30

“No early retirement”. This experiment on the environmental outcome of “promoting large

5

and closing small” policy quantifies the difference in term of the accumulated CO2 emission

from coal-fired power facilities with distinct age structures.

To answer this question, we need to make assumptions about what patterns of power

sector would look like without early retirement of small power units. As a straightforward

approach, the case without early retirement will be to meet the same amount of total coal-5

fired electricity demand (using the capacity as the proxy) in 2006-2013, and then keep the

old units online until their lifespan passed. As a result, the new additions in 2006-2013 and

subsequent years will be altered because of no retirement in advance and changed

depreciation pace.

Methodologically, this will need a vintage structure of the thermal power capacity, and the 10

assumption on the expected life time of coal power plants. The results will be intuitionally

sensitive to this setting, because it means the different turnover of power fleet. Here,

consistent with China’s convention on power plant design (China, 2011), we set the normal

life time as 30 years. Our measurement covers the period from 2005-2050, which is

sufficient to account for the cumulative difference due to the policy change (large-scale 15

early-retirement of small power units) in 2006-2013.

Our paper is structured as follows. In the next section, we explain the equations and data

processing regarding the vintage structure of historic coal fired units, addressing some data

problem through simplified assumption or calibration based on historic aggregated

statistics. Section 3 illustrates the construction of the scenarios in two dimensions, with 20

limitation on coal by 2020 or not, and early retirement (real case) or not (counter-factual

case). Section 4 presents results and its implication. Section 5 provides some discussion on

role of CCS, robust energy and environment policy, and the possibility of coal power

limitation policy choice in China.

2. Vintage capacity structure and calibration 25

2.1 Equations and data flow

The deprecation assumed here is 0/1 scrappage pattern, i.e. coal power units, with a normal

life time discarded when the life time is over, but before this, keep running. The total

capacity, early-retirement part, stock efficiency and emission of total fleet in Year t can be 30

expressed as:

6

=

>=

Where,

, the year in the analysis 5

, the capacity (GW)

, the new capacity additions (GW)

, the vintage of the corresponding capacity additions

, the early-retirement capacity (GW) part

, the energy efficiency (%) of capacity stock 10

, the energy efficiency (%) of new capacity additions

, the corresponding efficiency (%) of large-scale early retirement capacity

, Emission of coal-fired generations (billion tone CO2e)

, annual operating hours of the capacity (hours), set as 5000hours in the projection period and in the calculation of committed emissions. 15

, Emission factor of coal (tCO2/kWh), set as 2.8 tCO2/kWh

The data flow for the variable determination at the historic calibration (1970-2013) and simulation in projection period (2013-2050), and the period in “counter-factual” test (2006-2013) are illustrated as Fig.1. 20

7

[Figure 1 Schematic of simulation and calibration flow]

2.2 Historic stock and its efficiency

We have the historic stock, aggregated efficiency, and volume of early-retirement small 5

power units from official or semi-official statistics (Figure 2). This data foundation was

used for the calibration of vintage structure of power units from 1970-2013.

The capacity increased steadily over the entire period till 2013 (the base year for this

analysis) from 1970s. Divided by China’s specific Five-year Plan regime, despite periodic

economic growth and power sector development is crystal clear (Zhang and Liu, 2006), the 10

tight situation of electricity supply persisted through the whole period till 2005 and with

few exceptional years.

The severe electricity deficiency occurred in 1980-1990s encouraged the intensive

investment on power units from local government, energy-related ministries and

enterprises (Mou, 2014), and most units built in this period are small and less efficient. This 15

is the fundamental reason why from 1985 to around 2000, the efficiency of coal generation

improvement is very limited, only 3-4 points of percent.

8

After several years with extreme low energy (even negative) and power growth, from 2003,

another electricity supply deficiency occurred1, which brought another cycle construction of

small units, especially sponsored by the local government. These units are successively put

into operation in the coming 2-3 years, resulting in the similar slow efficiency improvement

till around 2006. 5

[Figure 2 Coal power capacity, early-retirement (ER) units, and the stock efficiency (1960-2013)]

Source: China power sector 50 years (1960-2000); China electric power yearbook (2000-2010); Power statistics flash report by CEC, China (2011-2013) 10

In 2004, NDRC of the government issued guidelines on the coal-fired power plant

construction, and regulating the units in principle should be above 600 MW, and adopt the

super- and ultra-super critical technology, and further, the coal consumption rate should be

better than 286 gce/kWh (i.e. about 43% efficiency) (NDRC, 2004). This policy, together 15

with the power sector reform (splitting grid and generation part, and introducing

competition between generators), brought a new round of large-scale, but high efficient

units into the power fleet since 2005.

At the same time, shutting down of old and small units was strongly launched. Figure

2(right chart) shows the annual capacity scrappage from 2006-2013, totally 77 GW coal-20

1 The detailed introduction and analysis on electricity deficiency is beyond the scope of this paper, which can be found at Zhang and Liu (2006) and Zeng (2013).

9

fired generation capacity. In some peak year like 2009, the closure part of the units could be

as much as 4.2% of the total coal power capacity.

This rapid size structural change from “promoting big, closing small” accelerated the

efficiency improvement. Since 2006, the stock efficiency increase by almost 1% annually as

the result. 5

2.2 Assumptions on the vintage-distribution of early-retirement (ER) units

The vintage of the early-retirement small plants is not available and the plant-by-plant

information is far from completeness to meet the total closed capacity level indicated in the

official statistical report. So we make an arbitrary assumption that the vintage of ER units 10

were distributed averagely spanning 10 years, and the newest units shut down at Year t was

the units built 20 years ago, i.e. with a vintage of t-20. If more comprehensive data sources

become available, this simplified assumption can be eliminated.

The year-by-year accounting of net additions and early-retirement part is given in a matrix

illustrated in Figure.3. In few years, remaining capacity (after normal depreciation) is not 15

sufficient numerically for the purpose of early retirement to satisfy the above setting on the

vintage distribution. In this case, the closure unit distribution in this year is set as ZERO and

this closing part is set to the vintage one year newer. This work is repeated until the

summation of early retirement vintage units matching the total ER volume, and the

remaining capacity for either of the vintage is greater or equal to ZERO. 20

10

[Figure 3 Schematic vintage structure of the power stock]

2.3 Efficiency of new additions

5

With the stock efficiency, capacity and the assumption on distribution of the closure part of

the small units, the efficiency of new additions in each vintage can be calibrated year by

year from the beginning year of our analysis.

For the first year 1970, we assumed that the total capacity was wholly added in this year,

and so stock efficiency is exactly equal to the efficiency of new additions. And this part will 10

last for the whole life time (i.e. 30 years later to the start of the year of 2000) if no early

retirement policy. This approximation in term of the vintage structure is still acceptable

given the capacity in 1970 was still very small compared to later years.

For the subsequent years, the new additions firstly are calibrated based on the total

capacity in the same year and the remaining capacity inherit from the previous year. And 15

secondly, the efficiency of this additional part was calibrated with the stock efficiency and

historical vintage capacity additions and their respective efficiency.

This work can be done from 1971-2013, and then a trajectory of efficiency of new additions

by vintage was obtained, shown at Figure 4.

11

[Figure 4 Efficiency of new additions by year]

Note: the efficiency level before 2014 is calibrated and afterward assumptions based on technical potential of cutting-edge technology IGCC units, 50% set as the max. level, and

achieved in the final year of our analysis. 5

3. Scenarios

For the efficiency of new additions in the projection period, exogenous assumption based on

scenario storyline is given. The efficiency of new additions are set to be increased quickly by

2020, and reaching the 48% as the theoretical maximizing potential of super-critical units,

then improve flatly to 50% by 2050. This is identical across scenarios. 10

[Table 2. Overview of scenarios]

Scenario Early Retirement No Early Retirement

No limitation on coal-fired power plants (BASE) BASE-ER BASE-noER

limitation on coal-fired power plants (POL) POL-ER POL-noER

Our scenarios cover two dimensions (Table 2): existence or not of the coal power limitation

policy, and with ER and assumed no ER. In case of coal power can increase to meet the

growing electricity demand in a free way, we need additional assumption on its volume in 15

total capacity for the sake of accounting cumulative emissions. We make the assumptions at

Table 3, which is roughly consistent with 2kw per capita in the year of 2030-2050, a

saturation level envisioned in many studies, e.g. NEA (2013). The share of coal is expected

to decline steadily along the fast penetration of advanced technologies, including nuclear,

wind and solar PV, from 63% in the total capacity to about 50% in 2030 and 40% in 2050. 20

20%

25%

30%

35%

40%

45%

50%

55%

1970 1980 1990 2000 2010 2020 2030 2040 2050

historical assumptions

12

[Table 3. Coal power generation capacity assumptions if no banning]

Scenario 2013 2020 2030 2050

Total capacity (GW) 1247 2000 3000 3100

Share of coal power capacity 63% 55% 50% 40%

Coal power capacity (GW) 787 1100 1500 1240

Assumption of new capacity of coal-fired plants banned or not will be the assumptions to

constitute the POL case. The reason for this ban, in realistic world, could be some direct

control, or regulation on the emission intensity to make coal to power technology and/or 5

economically infeasible (Eide et al., 2014 is an example in US case), or some other forms.

In another dimension, ER cases are the real case occurred at 2006-2013, and noER cases

assumed that no ER among this period. This is impacting on the new additions, and their

profile in the whole projection to 2050, to match the exogenous capacity trajectory, shared

within the BASE cases. In the POL cases, the new additions of coal power are ZERO from 10

2020 as the storyline.

The comparison between scenarios gives us implications on the impacts on CO2 emission of

ER and potential policy of limiting coal use.

4. Results

4.1 Different patterns of coal power fleet in 2013 15

The counterfactual test brought an alternative vintage structure of coal power sector

(Figure 5), which potentially influence the long term emission trajectory and cumulative

volume. Inspired by the work at Davis and Socolow (2014), the remaining emissions (the

total “committed” emission in the remaining life time if no early-retirement) in 2013 in the 20

ER and noER cases can be calculated.

The results shows that in ER case, the remaining emission from coal fleet is as much as 8.2

billion tone CO2e, contrast to 6.3% lower, i.e. 7.6 billion in noER case. Correspondly, the

early retirement results a younger age cohort in ER case. In 2013, the average age of the

13

whole coal power fleet is as old as 9.7 years in noER case, which was lifted to about 8 years

through early retirement in ER case instead.

[Figure 5 Different vintage structure in 2013 in ER and noER cases]

4.2 Disparity of total coal capacity 5

As the result of coal banning from 2020, the remaining coal capacity in the POL cases are

different between ER and noER cases. Figure 5 shows the magnitude of disparity.

Due to the banning from 2020, the built coal capacity in the power system phases out and so

the ER and noER cases are convergent to ZERO by 2050. But this process is distorted by the 10

difference in 2006-2013, causing different vintage structure and size of the remaining coal

capacity (Figure 5).

Compared to the noER cases, from 2020-2038, the ER case have a larger annual remaining

capacity about 7-37 GW accordingly (shadow area in the figure), arising from the effect of

younger fleet during this time horizon. The emissions from this part cannot be fully offset by 15

the more additions in noER case during the period of 2013-2020 (to meet the increasing

electricity demand), so bring a net emission increase effects for the whole period.

6 48

115

145 61

60

508 437

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

ER noER

After 2005

Before 2005, after 2000

Before 2000, after 1990

Before 1990

14

[Figure 6 Disparity in coal capacity]

4.3 Cumulative emission difference

Table 4 summarized the calculated emissions in the four scenarios, cumulative in several 5

periods.

It shows that early-retirement policy can bring about 0.4 billion tones of CO2 emission

reduction during 2005-2013, and further to 0.6 billion by 2020 due to higher efficiency

units more in the power fleet. But this mitigation effects is gradually weakened after 2020

due to the continuous cycle of “old plant phase-out, and new plant with high efficiency 10

entering the fleet”.

Large volume of coal power construction during 2005-2013 means that only after 2040s,

this part can be cleared out of the system. So compared to noER case with normal

depreciation, from the period of 2030-2050, this part negatively influence on the

performance of coal power again, and the cumulative emissions are 0.2 billion higher than 15

noER case. Totally speaking, the emission gap from policy intervention in 2005-2050 is as

small as 0.1 billion due to the unlimited cycle of coal power entering and steadily efficiency

improvement of new additions.

If there is coal banning in the power sector, the picture is significantly different after the

banning adoption, as the cases of POL-ER and POL-noER shows. All the small units built in 20

1980-1990s with a life time of 30 years, exit by 2020 if no early retirement, and after that

15

the remaining capacity of coal is smaller and smaller, due to no additions, and continuous

depreciation. But in the case of early-retirement, the new additions in 2006-2013 crowed

out the old ones, and persist for a longer time and bring more emissions. Cumulatively, the

ER policy brings 1.2 billion more emission during 2005-2050.

The emission reduction in the near term will be dimmed and fully offset by the increasing 5

“committed” emission in the remaining life time of the new facilities built at 2006-2013.

This is totally consistent with our intuition on the impacts of early-retirement policy if coal

power banned in the near future.

[Table 4. Cumulative emission in various scenarios]

Cumulative emission (billion tones)

Scenario 2005-2013 2005-2020 2005-2030 2030-2050 2005-2050

BASE-ER 26.8 56.5 110.8 115.8 220.7

BASE-noER 27.2 57.2 111.1 115.6 220.8

Delta (BASE) 0.4 0.6 0.3 -0.2 0.1

POL-ER 26.8 56.4 99.7 39.2 134.9

POL-noER 27.2 57.0 98.9 38.7 133.7

Delta (POL) 0.4 0.6 -0.9 -0.5 -1.2

10

4.3 Good or bad policy?

So based on the empirical measurement, how we can give a summary on the merit of the

policy “promote big and close small” adopted in China from 2006-2013. There is no a simple

answer to say this is a good, or bad policy given there is no single prescriptive standard. 15

Adopting the common criteria to assess the environmental policies (Aldy et al, 2003;

Gouder and Parry, 2008), Table 3 outlined the performance of early-retirement of small

power units in China against these criteria, from this study and others, including

quantitative and qualitative studies.

Six criteria can guide an assessment of the policy adopted from 2006-2013: (1) the 20

environmental outcome; (2) cost-effectiveness; (3) dynamic efficiency; (4) distributional

equity (both cross-sectional and intertemporal); (5) flexibility in the presence of new

information; and (6) participation; (7) compliance.

16

Our paper based on the vintage structure is an assessment on the greenhouse gas related

environmental outcome. Generally, long-term positive environmental effect is not assured,

especially for global polluter long-lived CO2 emission with stock nature.

And the criteria of cost-effectiveness can be shown by the marginal abatement cost, and

compared with other mitigation options (e.g. efficiency improvement, demand saving, and 5

renewable and nuclear switch) in power sector and beyond. From Climate Institute (2010)

and our calculation, the marginal cost of ER policy is around 800Yuan/tCO2e, a level beyond

CCS even. From this meaning, ER policy is “expensive” and inferior, whatever in the short

term or long term.

In term of dynamic efficiency measurement, i.e. whether the benefit as a result of the policy 10

outweighs the associated cost, many additions things need to be addressed. The benefit of

ER policy is difficult to quantify, given there is no clear accounting boundary especially

accounting for local pollutions. The revenue on energy security and health revenue

complicates the questions with substantial uncertainty, which is another broader topic

beyond the scope of this paper. But in the long run, even the positive environmental 15

outcome cannot be assured (as well the benefit from alleviated local damage if emission

standards not tightened in the future), dynamics efficiency is unlikely in a qualitative

estimation.

The estimation of distributional equity issue here focus on within the coal power sector. The

criterion is the size and age of the units, to determine whether the small power plants being 20

listed in the closure category or not. For the sake of employment and social security

concerns, the owner can normally get some compensation or subsidy in the short run. In the

long run, a larger unit can be built with the “allowance” of the previous small one, and this

right could be tradable as well (Price, et al, 2011; NDRC, 2007). With these additional

measures, “the loser” from the policy, i.e. the owners and employers of the closed plants, 25

could be better-off, or at least no worse-off.

Participation is limited to the small powers on the list as this is command-and-control policy,

and nothing relevant to other peer plants in power system. Compliance is achieved fully

through strong political task distribution and monitoring, taking this policy as an

emphasized priority of local government and respective ministries. 30

17

Flexibility in the presence of new information is impossible for this rigid policy (the closed

power plants dismantled). Of course, in short term, no significant policy change expected

and it is robust. But from an intertemporal perspective dealing with very long time horizon

climate change problems, lack of flexibility of this policy is in sight.

5 [Table 3. Performance of early-retirement of small power units in China]

Short term (2006-2013) Long term (2006-2050)

Environmental outcome

The thermal power sector mitigated CO2 300 million tones in 2010, relative to a constant efficiency benchmark of 2005. (CEC, 2011)

If no policy limitation on coal power, by 2050, the policy bring cumulative emission reduction around 10 million tone.

Cumulative emission from 2005-2013 is reduced by 270 million based on vintage structure in this paper

If coal power banned from 2020, the policy bring cumulative emission increase around 190 million tone during 2005-2050.

Cost-effectiveness

Marginal abatement cost about 800 Yuan/tCO2 (XXX, 2010; Climate Institute, 2010) due to the comparison between total cost of new plant and operating cost of old plant, which is beyond CCS.

Similar to the short term case.

Dynamic efficiency

The benefit from reduced energy consumption and emission, energy security and health revenue need further assessment

If coal power banned from 2020 existing, not likely if the local polluter standard not improved in the future. In other cases, likely.

Distributional equity

Likely for the owner of the small units after getting some subsidy or compensation for the sake of employment and social security.

Yes, because generally the owner of previous small units, as the potential “loser”, can be compensated by launching a larger unit or trading such a right to others.

Participation No. Only the power plants listed in the scope involved

Not applied.

Compliance Yes as emphasized task of local government and ministries.

Short-term “command-and-control” policy, not applied.

Flexibility in the presence of new information

No. But short term, no significant policy change expected.

No. This paper showed this point exactly.

18

5. Discussions

5.1 CCS not game changer

Not surprising, different assumptions can generate very different answers. CCS is one 5

technology option to reduce the carbon intensity of coal power, to capture 90% and even

higher emissions. In this analysis, we exclude the possibility of CCS equipped with the coal-

fired power plants, as an effective measure to hedge the fortune of coal in stringent climate

constraint. This is an assumption for the sake of simplify, and low expectation on the

prospect of CCS. 10

Many latest studies verified this point. Strand and Sebastian (2014) developed a two-staged

stylized model to simulate the impacts of long-lived infrastructure and consider the options

including retrofit, costly CCS or abandoning, and found infrastructure investments could be

made without sufficient concern for future climate costs, but these costs still actually

incurred when the future arrives. Eide et al., (2014) considered the capture rate in a 15

continuous range and they found that shift from coal to gas, rather than investment on CCS

is more likely change with the proposed levels of emission standards. Zhang et .al (2014)

simulate the efficient power mix under the uncertainty of various technology performance,

and showed that the time window of CCS deployment in coal amenity is very narrow and

temporarily, and inferior to the improved renewable technologies. Fuss et al. (2004) 20

showed that only CCS equipped with biomass to be negative emission is meaningful for 2

degree target scenarios.

If no technology breakthrough, it is hard to image that CCS will be a game changer in tight

climate constraint world, especially combined with most carbon intensive coal.

5.2 Envision of the possibility of coal limitation from 2020 25

The energy situation in China in the past years changes little and gradual, but the massive

problem of environment degradation, especially air pollution is becoming central concern.

China’s fight against the local air pollution has been launched since 2013, stimulated by the

widely-known air pollution episode that occurred in the beginning of 2013 and covered 30

one-sixth of China’s territory. As an important component, to reduce the “dirty” coal use and

19

change the utilization way was pledged in government planning (GOC, 2013) to address the

problem.

At the same time, the signal of power overcapacity in China is seen from the supply and

demand pattern of electricity in 2014. Negative growth in August, and only 4% growth for

the whole year is expected (Dale, 2014). Towards a longer horizon to 2020, the slowdown 5

of economic growth and the strong competition of other power options, would make the

coal-fired generation approval2 and construction halt, temporarily or permanently with a

more and more vast possibility.

This possibility is strengthened with the consideration of climate mitigation. People can still

argue that coal endowment in China is abundant and phase-out of coal utilization 10

(especially coal power) is costly, and the measures in non-power coal use sectors (e.g.

households, iron & steel) and improve the emission standards might be more cost-effective.

But this argument can’t be justified adding the climate constraints, upon power sector is the

most key sectors for climate mitigation. To meet the 2-degree target to secure the climate

system, about two-thirds of proven fossil-fuel reserves must remain in the ground, mostly 15

coal (IEA, 2013). Reached a firm commitment and target in Paris climate meeting, the coal in

power sector, characterized with large committed emission as long-lived infrastructure, is

difficult to find its role in power sector towards deep decarbonization.

6. Conclusion

The large-scale coal-fired power plant substitution with large size units for small units in 20

China from 2006-2013 lifted China’s current power efficiency in a quick way, and will

impact on the pace and profile of coal power in China in the next 20-30 years, even further.

In this paper, we measure the dynamic environment outcome of the policy, specifically on

the co2 trajectory of coal power under new additions of coal possible or not.

Retroperspectively, it shows that whether this large-scale closure of old plants increase or 25

decrease the long term emission depend on the climate policy choice and its impacts on the

coal. If by 2020, the new addition of coal power is forbidden, the cumulative emission in

2 In China, the construction of commercial coal-fired power plants is administrated by the central government, and its scale and layout are still controlled by National Development and Reform Commission (NDRC) in the form of power sector planning or other regulations.

20

2005-2050 in coal power sector is 1.2 billion tones more than the counter-factual case, i.e.

this part normally decommissions after the lifespan finished. This implies “stranded” risk of

large-scale coal power asset because of the possible new climate policy and its uncertainty.

Long-term positive climate effect is not assured, unless coal utilization equipped with costly

carbon capturing and storage (CCS), which would possibly make coal-fired generation 5

economically unviable relative to the renewable and other options with learning potential.

The cost of implementing future climate mitigation policy might be altered as well.

Energy infrastructure is characterized by long life time, which means a strong inertia and

carbon lock-in path dependence once put into operation. Long-term consequences of

current energy investment and policy choices should be considered explicitly in the context 10

of plausible climate mitigation policy.

References

Zhang, Shuwei, and Nico Bauer. "Utilization of the non-fossil fuel target and its implications in China." Climate Policy 13.3 (2013): 328-344.

National Energy Administration (NEA), 2014. Annual energy work meeting of China 2014, available 15 at http://www.nea.gov.cn/2014-02/11/c_133105714.htm.

Ouyang Chancy, 2014. Personal communication at May 15, 2014.

Price L., D.M. Levine, N. Zhou, D. Fridley, N. Aden, H.Y. Lu, M. McNeil, N.N. Zheng, Y.N. Qin, P. Yowargan. Assessment of China's energy-saving and emission-reduction accomplishments and opportunities during the 11th Five Year Plan. Energy Policy, 39 (4) (2011), pp. 2165–2178 20

China Electricity Council (CEC), 2011. China power mitigation 2011. Available at http://huanzi.cec.org.cn/dongtai/2011-11-22/74594.html

Wang, Qiang, and Yong Chen. Energy saving and emission reduction revolutionizing China's environmental protection. Renewable and Sustainable Energy Reviews 14.1 (2010): 535-539.

Davis S J, Caldeira K and Matthews H D, 2010. Future Co2 emissions and climate change from existing 25 energy infrastructure Science 329 1330

Davis, S. J., and R. H. Socolow. "Commitment accounting for CO2 emissions." Environmental Research Letters, 9 (2014) 084018.

China, 2011. Code for design of medium- to large-scale fossil fired power plants GB50660. Available at http://www.codeofchina.com/gb/architectural/12484.html 30

National Development and Reform Commission (NDRC), 《国家发展改革委关于燃煤电站项目规划和

建设有关要求的通知》，发改能源[2004]864 号文。Available at http://www.nea.gov.cn/2012-

01/04/c_131262602.htm

Zhang S, and Liu D, 2006. Analysis of the relationship between power shortage and economic growth, Power demand side management, 8(1), 11-13 35

21

Dunguo Mou, 2014. Understanding China’s electricity market reform from the perspective of the coal-fired power disparity, Energy Policy, in press. http://dx.doi.org/10.1016/j.enpol.2014.09.002.

Zeng Ming, Xue Song, Li Lingyun, Wang Yuejin, Wei Yang, Li Ying, China's large-scale power shortages of 2004 and 2011 after the electricity market reforms of 2002: Explanations and differences, Energy Policy, Volume 61, October 2013, Pages 610-618 http://dx.doi.org/10.1016/j.enpol.2013.06.116. 5

National Energy Administration (NEA), 2013. Medium to Long term generation capability and electricity demand projection. Available at http://www.nea.gov.cn/2013-02/20/c_132180424_2.htm

Eide, J., de Sisternes, F.J., Herzog, H.J., Webster, M.D., 2014. CO2 emission standards and investment in carbon capture. Energy Economics, 45:53-65

Government of China (GOC). Air Pollution Prevention and Control Action Plan of China. 10 <http://www.gov.cn/zwgk/2013-09/12/content_2486773.htm> (in Chinese), Sept. 10, 2013 (2013)

LAWRENCE H. GOULDER, Ian W. H. Parry. “Instrument Choice in Environmental Policy”, Review of Environmental Economics and Policy 2(2), 2008

The Climate Institute, 2010. The implicit price of carbon in the electricity sector of six major economies. Available at http://www.vivideconomics.com/uploads/reports/the-implicit-price-of-15 carbon-in-the-electricity-sector/Vivid_Econ_Implicit_Carbon_Prices.pdf

Shuwei Zhang, Nico Bauer, Gunner Luderer, Elmar Kriegler. Role of technologies in energy-related CO2 mitigation in China within a climate-protection world: a scenario analysis using ReMIND, Applied Energy, 2014, 115: 445-455.

Fuss et al. (2014). Betting on Negative Emissions. Nature Climate Change (commentary) 20

Strand Jon, Miller Sebastian, Long-Run Carbon Emissions Implications of Energy-Intensive Infrastructure Investments with a Retrofit Option, Energy Economics (2014), doi: 10.1016/j.eneco.2014.10.002

Dale Katie, 2014. China’s thermal power output recorded a downfall. http://marketrealist.com/2014/10/chinas-thermal-power-output-recorded-downfall/ 25

International Energy Agency (IEA), 2013. World Energy Outlook2013. OECD/IEA.

NDRC (2007). Inform regarding accelerating the closure of small power units, available at http://www.gov.cn/zwgk/2007-01/26/content_509911.htm.

30

Table

[Table 1. Parameterization of the learning]

[Table 2. Overview of scenarios]

Figures

[Figure 1. Emission of energy-related CO2 in BASE & POL scenarios] 35

22

[Figure 2. Global average investment cost of wind and solar PV and generation expansion in

POL scenarios]

[Figure 3. Interaction of generation expansion and learning in major economies for wind

power]

[Figure 4. GDP accumulated loss (2010-2050) in policy scenarios.] 5

[Figure 5. Emission of electricity CO2 with changed formula of learning]

[Figure 6. Wind and solar PV generations with changed formula of learning]