A system dynamics analysis of technology, cost and policy that affectthe market competition of shale gas in China

Wu Yunna, Chen Kaifeng, Yang Yisheng n, Feng TiantianSchool of Economics and Management, North China Electric Power University, 102206 Beijing, China

a r t i c l e i n f o

Article history:Received 23 April 2014Received in revised form12 November 2014Accepted 19 January 2015Available online 9 February 2015

Keywords:Shale gasChinaSystem dynamics modelCompetition

a b s t r a c t

In order to ease environment pressure and alleviate the scarce problem of energy, Chinese government hasbegun to exploit and develop shale gas (SG) since 2009. However, the formation of competitive market isbristled with difculties, due to its infancy, imperfect and lack of standardization. This paper rstly analyzes thestatus quo of technology, policy, cost, and competition of SG industry in China. Then a system dynamics (SD)model is built to show various trends of Chinese SG industry under different scenarios, which shows thattechnology, policy and cost all have effects on competitiveness. At last, the simulations reveal that the numberof competitors in Chinese SG industry will arrive at its peak from 2019 to 2020. The correspondingrecommendations have been presented: Chinese government should pay more attention on the perfectionof laws and regulations of SG industry. And competitors should focus on increasing R&D investment.

& 2015 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2362. System dynamics method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2363. System dynamics model of SG industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236

3.1. Structure analysis of competition system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2363.2. Model assumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2373.3. Variables and parameters of SD model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237

4. Case study and data collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2394.1. Exploitation cost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239

4.1.1. Cost estimation of drilling process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2394.1.2. Cost estimation of well completion process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2394.1.3. Cost estimation of well cementation process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2394.1.4. Cost estimation of logging process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239

4.2. Government policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2394.3. Competition situation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240

5. Simulation results analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2415.1. Simulation software. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2415.2. Result analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241

5.2.1. Technology scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2415.2.2. Cost scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2415.2.3. Subsidy scenarios. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2425.2.4. Combined inuence of technology, cost and policy on market competition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242

6. Conclusions and suggestions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243

Contents lists available at ScienceDirect

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

Renewable and Sustainable Energy Reviews

http://dx.doi.org/10.1016/j.rser.2015.01.0601364-0321/& 2015 Elsevier Ltd. All rights reserved.

n Corresponding author. Tel.:86 13401067697.E-mail address: 8629823yys@163.com (Y. Yisheng).

Renewable and Sustainable Energy Reviews 45 (2015) 235243

1. Introduction

As a coal-dominated energy structure country, China is suffer-ing from serious environmental problems. It has become asignicant energy strategy for China to develop clean energy.Natural gas, a type of clean energy, is becoming increasinglypopular. Gas consumption increases year after year. However,China's conventional natural gas reserves fail to meet the increas-ing demand of gas. Therefore, reliance on imported natural gasposes a great energy security threat to China. More than 31.6% ofthe gas consumption relied on foreign supplies according tostatistics in 2013 [1]. Energy security in China will become moreserious over time. Fortunately, China has huge reserves of shalegas (SG) which is a type of unconventional natural gas [24]. It willbe expected to effectively alleviate above problems when SGbegins to be commercially exploited to a certain extent. Manyscholars having studied the SG industry believed that SG devel-opment is an energy revolution which will have a great inuenceon the world energy landscape [57]. Many regions around theworld like European countries, Australia and China begin toresearch or exploit SG after the US.

Chinese SG industry has been in the forming stage since 2009and SG market is easily monopolized, which is unfavorable fortechnology progress, cost reduction and market competition.Competition benets to realize efcient use of energy resourcesand low extraction cost. Therefore, healthy and competitive SGmarket is particularly important to China energy strategy. In fact,SG market is affected by complex technology, cost and policyfactors in China because of deeper burial depth and more complexgeological conditions [811] and further, these factors have sig-nicant inuence on the competition of SG market [12]. Somescholars qualitatively listed several key factors affecting the SGindustry competition: technology, cost and policy [1319]. Hu andXu [15] advocated that the increase of researches on SG exploita-tion and government subsidy might benet SG industry develop-ment. However, the inuence mechanism between these factorsand their inuence on SG market is lack of quantitative research indetail. This paper aims to analyze these factors affecting thecompetition of SG market and corresponding inuence on com-petition. Qualitative and quantitative methods are applied toensure comprehensive and sufcient argumentation.

We try to choose a system-analysis model that can analyzecomplex system to study the above issues. Estimating the inuen-cing mechanism in an industrial section is a complex problem dueto the presence of multiple technological limitations, feedbackprocesses among subsystems, and various kinds of delays. Systemdynamics is a suitable approach to such a complex model. Systemdynamics (SD) approach, established by Forrester [20], has beenwildly used in socio-economic studies. Many researchers developedSD models to analyze industry in various districts all over the world[2224]. SD also has wide application in energy eld. Feng, Chenand Zhang developed a model to study the energy consumptionand CO2 emission trends of Beijing from 2005 to 2030 [25]. Naillpresented an SD model of gas industry dynamics in the UnitedStates. He demonstrated technological, physical, economic andpolitical factors that might alter the gas industry [26]. Movilla,Miguel and Blzquez[27] used SD to analyze the behaviors of thephotovoltaic sector in Spain and its expectations in some possiblescenarios. Ansari and Sei studied energy consumption and CO2emission in Iranian cement industry under various production andexport scenarios based on SD model [28]. Kiani and Ali presented asystem dynamic model which considers the feedback betweensupply and demand and oil revenue of the existing system in Iranconsidering different sectors of the economy [29].Chyong, Nuttalland Reiner presented an SD model of the indigenous natural gasindustry in the UK [24]. Therefore, advantages of using SD model to

analyze the inuences of complicated factors on industry competi-tion are obvious from the published literatures mentioned above.

In conclusion, it is very important to form a healthy market inan orderly and efcient way because of the strategic importance ofSG. Through literature comparison, we nd that SD model is afeasible method to analyze the competition of SG industry. In thispaper, we build the SD model and analyze the inuence oftechnology, cost and policy on the SG industry. The results andconclusions will be helpful for government to formulate policies,and for enterprises to make decisions.

2. System dynamics method

SD method was originally developed to help corporate man-agers to improve the understanding of industrial processes [20].SD focuses on describing endogenous feedback structure of asystem. The word endogenous means growing from a system.An endogenous structure generates the dynamics of a systemthrough the interaction of variables and actors represented in themodel [30]. It is an important method to learn and solve problemsof complex system. It emphasizes to take a view angle to treat theresearch issues systematically and objectively.

Through modeling feedback structure, one can better under-stand dynamic complexity that can be found in situations wherethe same action has different effects in the short and the long run,and where an action has one set of consequences locally and adifferent set of consequences in another part of the system. It hasbeen argued that real leverage in many management and economysituations lies in understanding dynamic complexity. Systemdynamics has been widely utilized to study dynamic behaviorsof various social systems, and has been applied in policy formula-tion and analysis both in the public and private sectors.

To be specic, the SD model mainly includes the following steps:

1) Analyzing system structure to understand elements of thesystem and relationships among them. This step allowsresearchers to get a clear boundary of a system and nd outthe systematic characteristics and structures macroscopically.

2) Establishing ow diagram to further describe logical structureof the system. The modeling elements can be divided into levelvariables, rate variables and auxiliary variables. They reect thenature of the analyzed object correctly with standard symbols.

3) Constructing the equation for these variables in a ow diagramto analyze the quantitative relationships between them clearly.The essence of SD equations is taking a differencing treatmenttowards a series of differential equations. On the one hand, itdescribes the relationships between the variables in mathema-tical form. On the other hand, it also reects the recursiverelationships between the variables. The construction of equa-tions makes it easy to analyze a complex systemwith computersimulation.

4) Entering these ow charts and equations into computer simu-lation environment to simulate, debug, and examine the modelso that we can adjust control variables in the model to studythe inuences of different factors on the system.

3. System dynamics model of SG industry

3.1. Structure analysis of competition system

The exploitation of SG calls for high level of technology. Whentechnology barriers decline or even disappear, market competitionmay become erce for the reason that more companies acquiring

W. Yunna et al. / Renewable and Sustainable Energy Reviews 45 (2015) 235243236

exploitation technology will enter into this SG eld. Moreover thecomplex exploit technology leads to the high cost directly. In theinitial stage of market formation, companies have limited protdue to high cost. Market competition will grow intensied, whenthe cost comes down. The reason is that more companies willchoose to enter into SG market, once these companies can gainmore prot than expected. Given SG resource belonging to Chinesegovernment, obtaining the exploit right is closely related withgovernment policy. So far, tendering and bidding meetings of SGexploit right were held by Ministry of Land and Resources of thePeople's Republic of China (MLR) in 2011 and 2012. The third timetendering and bidding were not held in 2013 for some reasons.The types of the bid-winning enterprises changed tremendously,due to the changing policies. Private enterprises have beenallowed to enter into the SG industry since 2012, which meansSG market competition intensies thanking to lower barriersto entry.

Based on the backgrounds and data above, the structurecharacteristics of SG market are obtained and listed as follows:

(1) The level of technology inuences the production, cost, policyand competition. First of all, complex technology requirementswill form barriers, which may affect the market competition.Then, the improvement of technology can increase gas pro-duction and decrease its cost so that the prot becomesgreater and more competitors will be attracted to enter intothis industry. Furthermore, when technology lags behind,various supportive policies would be put forward by thegovernment to stimulate the enterprise investment andresearch behaviors. When technology reaches a relativelyadvanced level, the government is likely to repeal supportivepolicies.

(2) Cost affects SG industry competition. Cost plays a critical rolein enterprises' prot and the speed of capital recovery directly,which further inuences the decision-making of enterprises.Once the cost decreases, more enterprises will enter into SGindustry. Therefore, the cost affects the prots directly andaffects the market competition indirectly.

(3) Production has an inuence on SG industry competition.Besides, it also determines the prot and the speed of capitalrecovery. At the same time, the subsidies will depend on gasproduction according to the Chinese subsidy policies. There-fore, the production affects the prots directly and affects themarket competition indirectly.

(4) Policy makes differences in the technological level and com-petition. The market competition is affected by policy largelywhen the SG industry has not yet fully evolved. Entry and quitpolicy have affected the number of industrial competitors.Investment risk and prot margins can be uctuated by theimpact of subsidy policy and that also applies to the number ofpotential competitors.

Based on the above analysis, we can build the qualitative SDconceptual model of SG industry competition as shown in Fig. 1.mainly reects the inuence ways of different factors on thecompetition intensity. The arrows in Fig. 1 represent the inuentialrelationship between two factors. The positive sign and negativesign represent the positive and negative correlations, respectively.The inuence mechanism of technology, cost, policy and gasproduction are considered in the SD model.

3.2. Model assumption

This study makes following basic assumptions in order toquantify the structure model and build complete system dynamicsmodel.

Assumption 1. Competitive intensity is represented by bothenterprises in the industry and potential enterprises who prepareto enter into the industry.

Assumption 2. Technological level is represented by comprehen-sive stock of technology which is expressed with accumulativetotal investment.

Assumption 3. Cost will tend towards stability, when the stock oftechnology reaches a certain level. Each extraction cost of shalegas will decrease linearly with technology improving.

Assumption 4. Policy factors mainly include entry policy, exitpolicy, R&D policy and subsidy policy.

3.3. Variables and parameters of SD model

SG competition market is a multi-stage feedback systemwith timedelay and nonlinear characteristics. SD model can better describe theinuence and feedback mechanisms among multiple subjects incomplex system, so it is more suitable for analyzing SG market. Theinternal subsystem of SG industry competition is analyzed in details tobuild up the complete SD model. All parameters and their connota-tions are claimed in Table 1. Fig. 2 shows the complete SD model. Theblack, blue, pink and caramel parameters are technology, cost, policyand competition parameters respectively.

(1) Policy parametersPolicy parameters include four parts: entry policy (ETP), exit policy(EP), R&D policy (GFR&D) and subsidy policy (SP). ETP represents aset of random values to reect the number of enterprises that areallowed to enter into this industry by government. Exit policy (EP)has two kinds of situations inwhich the enterprises are allowed ornot allowed to exit expressed with 1 and 0 in the SD model,respectively. EP can affect the number of competitors that exit SGindustry every year (NEC). Subsidy policy (SP), gas production rate(GPR) and gas price (P) can directly affect enterprise's actualincome. Both government funding of R&D (IR&D) and EnterpriseR&D investment (GFR&D) inuence the stock of technologicallevel (ST). This paper uses the conditional statement to expressthis inuence shown in the following formula:

IF THEN ELSESTo\\! X; COSTGMC n R&D= XY ; 0 1Where X is the threshold of ST when the cost gets stable; COSTrepresents the cost of base period; GMC is the target cost; Yrepresents stock of technology of base period.

(2) Technology and cost parametersTechnology can have a signicant impact on the cost of shale gasdevelopment. We assume that advances in technology cangradually lower the cost of technology development. The connec-tions of technology and cost can be represented by the following

Fig. 1. SD conceptual model of SG industry competition.

W. Yunna et al. / Renewable and Sustainable Energy Reviews 45 (2015) 235243 237

parameters: MCRR, DFCRR, DLCRR and FCRR which can inuenceMC, DFC, DLC and FC, respectively. However, the cost will not fallwithout limit and it will remain stabilized around an estimate, likeGMC, GDFC, GLWDC and GFC. Each part cost will be summarizedas the total cost of the SG development (ACD). ACD and income ofenterprises (AR) affect the prot level together. Furthermore,prots inuence enterprises to enter or exit this industry. The

above parameters satisfy the following relations.

PR ARACD =ACD 2

ACD CTCCLCDLCLCMCOC 3

FC COSTFCFCRR 4

MC COSTMCMCRR 5

DFC COSTDFCDFCRR 6

LWDC COSTLWDCLWDCRR 7

DLCDFCLWDCODC: 8

In addition, this paper considers that the advanced technologyfrom research to application takes a certain amount of time, so adelay function is set in the model

TCDELAY1I ST; 0:5; 0 : 9

(3) Competitive intensityCompetitive intensity is represented by the number of compe-titors in the industry (LCP) and the number of potentialcompetitors (NPC). The calculation formula of LCP is as follows:

LCP CNI=NPCLCP: 10

NPC meets the probability distribution of PR, which meansdifferent PR decides the number of NPC. Therefore, a LOCKUPfunction in Vensim PLE is used to t their relationship.

Table 1Parameter list of SD model of SG industry competition.

Parameters Connotation Unit

Level variables ST (stock of technology) Technological level of SG industry (Ten thousand CNY)AR (average revenue) Annual revenues (Ten thousand CNY/Year)LCP (level of competition) Competition intensity

Rate variables R&D (total R&D investment) R&D investment (Ten thousand CNY/Year)GPR (gas production rate) Gas production rate (Ten thousand m3/Year)CNI (competitors number in industry) Competitors number in industry (Unit/Year)

Auxiliaryvariables

NPC (number of potential competitors) Number of potential competitors (Unit/Year)NC (number of competitors) Number of competitors that entry SG industry every

year(Unit/Year)

NEC (number of exit competitors) Number of competitors that exit SG industry every year (Unit/Year)LWDC (logging while drilling cost) Logging while drilling cost (Ten thousand CNY/Well)MC (measuring cost) Measuring cost (Ten thousand CNY/Well)DFC (drilling uid cost) Drilling uid cost (Ten thousand CNY/Well)DLC (drilling cost) Drilling cost (Ten thousand CNY/Well)FC (fracturing cost) Fracturing cost (Ten thousand CNY/Well)MCRR, DFCRR, DLCRR, FCRR (MC, DFC, DLC, FC reducingrate)

MC, DFC, DLC, FC reducing rate (Ten thousand CNY/Well/Year)

ACD (average cost in domestic) Average cost in domestic (Ten thousand CNY)CLC (completion cost) Well completion cost (Ten thousand CNY/Well)PR (prot rate) prot rate (Exclude tax) (Ten thousand CNY)TC (technology conversion) Technology conversion (Ten thousand CNY)

External variables IR&D (enterprise R&D investment) Enterprise R&D investment (Ten thousand CNY/Year)GFR&D (government funding of R&D) Government funding of R&D (Ten thousand CNY/Year)SP (subsidies of government) Subsidies of government (CNY/m3)ETP (entry policy) Entry policy EP (exit policy) Quit and exit policy CTC (cementing cost) Well cementation cost (Ten thousand CNY/Year)PC (perforating cost) Perforating cost (Ten thousand CNY/Year)LC (logging cost) Logging cost (Ten thousand CNY/Year)ODC (other drilling cost) Other drilling cost (Ten thousand CNY/Year)OC (other cost) Other cost (Ten thousand CNY/Year)P (price) Gas price (CNY/m3)GLWDC (goal of logging while drilling cost) Goal of logging while drilling cost (Ten thousand CNY/Year)GMC (goal of measuring cost) Goal of measuring cost (Ten thousand CNY/Year)GDFC (goal of drilling uid cost) Goal of drilling uid cost (Ten thousand CNY/Year)GFC (goal of fracturing cost) Goal of fracturing cost (Ten thousand CNY/Year)

Fig. 2. SD model of SG industry competition.

W. Yunna et al. / Renewable and Sustainable Energy Reviews 45 (2015) 235243238

We study this industry for the next 10 years and divide it intothree stages articially: the early stage from 2013 to 2016, themedium stage from 2017 to 2019, the last stage from 2020 to 2023.

4. Case study and data collection

Based on the SD model above, this study collects relevant datain China which will be as the input values of the SD model. Thedata mainly includes cost estimation, government policy andmarket competition.

4.1. Exploitation cost

This section states the cost estimation of China SG in 2013 andits change trends in future [31,32]. The data comes from Sinopecby 2013.

4.1.1. Cost estimation of drilling process

(1) Logging while drillingLogging while drilling (LWD) technology is the dominatingtechnology paradigm in China. LWD is divided into low-endproduct and high-end product. The cost of low-end productand high-end product are about 3 million CNY (485 thousand USdollars) and 15 million CNY (809 thousand US dollars), respec-tively. More enterprises may use the matured technology byconsidering the cost because low-end LWD technology is rela-tively mature in China. So this model assumes that the cost ofLWD keeps stable in 3 million CNY (485 thousand US dollars).

(2) Drilling uidDrilling uid should be prepared and be compounded accord-ing to different geological conditions. The core preparingmethod has been controlled by only a few internationalcompanies, like Halliburton, Baker Hughes and Schlumberger.These companies sell recipes of drilling uid to exploitationenterprises at the price of 68 million CNY (971 thousand to1.29 million US dollars). Sinopec has been researching drillinguid but its technologies are still not yet rened. In order toreduce cost, one recipe of drilling uid is used in similargeological conditions by some Chinese companies. It isexpected to reduce its cost to 2 million CNY (323 thousandUS dollars) when the core technology is developed [31].

(3) Drilling constructionThe cost of drilling construction can be estimated throughanalyzing the cost of drilling teams. The cost of personnel andequipment will take a total of about 25 thousand CNY/day(4.04 thousand US dollars/day) when the drilling teams haveno work during contract period. The cost is about 10002000 CNY/m (161.8323.6 US dollars/m) when drilling teamsare in working days. The vertical depth of SG in China is about35006000 m in China, the bending length is about 1000 mand the horizontal length is about 1500 m. The fastest speed todrill well is about 1000 m/day. So it will take 60 days to nishthe drilling tasks including 51 working days and 9 rest days.The cost of drilling construction is about 18.275 million CNY/well (2.95 million US dollars/well) based on the above givenconditions. The cost of drilling construction is relatively stable,so we assume it to be the constant in our SD model.

4.1.2. Cost estimation of well completion processThe core technologies of well completion are perforating and

fracturing. Sinopec has mastered the relative perforating technol-ogy which has reached the world-advanced level. The price ofperforating is about 2 million CNY (323 thousand US dollars).

China can produce the advanced large fracture devices such asSYL3000 fracturing pump, SHS20 mixed sand device, SYQ3000control device and auxiliary device. The cost of fracture service isabout 215 million CNY (323 thousand US dollars to 2.42 millionUS dollars) at current technical level. The well completion cost of ahorizontal well requires a maximum of 17 million CNY (2.75million US dollars). However, the total cost of well completioncan be declined to 4 million CNY (637 thousand US dollars)because there is large compression space in the cost of fracture.

4.1.3. Cost estimation of well cementation processWell cementation technology is already relatively mature. So

we assume it as a constant variable because of its relatively stablecost. The raw material cost and labor cost are the main costs. It isabout 5 million CNY (809 thousand US dollars) to cement one well.

4.1.4. Cost estimation of logging processLogging is a two-stage process: the rst process and second

process. The second stage also called digital logging. They havedifferent costs.

The technology of the rst logging process is also relativelymature. Domestic cost of logging is about 160200 thousand CNY/month (25.932.2 thousand US dollars). Logging period is about 60days, similar to the drilling cycle. The largest cost of logging isabout 400 thousand CNY (64.7 thousand US dollars) which isrelatively stable. So this article assumes it as a constant variable.

The cost of digital logging system is about 20 million CNY (3.2million US dollars). The cost of domestic image logging system isabout 50 million CNY (8.1 million US dollars) and the importedone is about 100 million CNY (16.1 million US dollars). Labor andequipment maintenance costs about 1.5 million CNY (242.7thousand US dollars) in its whole life. This equipment can be usedabout 10 years, and can measure about fty wells every year.Therefore, the average logging cost of each well is 260 thousandCNY (42.1 thousand US dollars), and the total logging cost of eachwell is 320 thousand CNY (51.8 thousand US dollars). It is expectedto reduce its cost to 200 thousand CNY (32.3 thousand US dollars)when this technology is developed.

4.2. Government policy

(1) Entry policyChina's SG exploration has been monopolized by state-ownedenterprises before 2012. Six state-owned enterprises areinvited to bid four blocks which own rich SG resources andtwo of them won the rights. In order to attract more capital,Chinese government allows private companies and other typesof enterprises to enter into the industry after 2012. Amongthese, two private companies and eighteen state-owned com-panies won the blocks.

(2) Quit policyQuit policy is a kind of method to reduce the investment risk.However, it is not formed in Chinese SG industry. Chinesegovernment requires the bid-winning enterprises to continueinvesting exploitation of SG according to the contract scheduleduring the contract period. If the enterprise cannot completethe tasks required by contracts, Chinese government mayrecover the exploration rights, which no doubt increasesinvestment risk of enterprises. The United States has formeda relatively perfect quit mechanism which creates benecialconditions to reduce investment risk.

(3) Subsidy policyLike coal-bed methane (CBM) subsidies policies, the SG enter-prises also can receive subsidies from the central govern-ment and local government. Government subsidy is a kind of

W. Yunna et al. / Renewable and Sustainable Energy Reviews 45 (2015) 235243 239

compensation for enterprise supported by state subsidies forspecic political, economic and social purposes. Subsidy ismainly provided to enterprises that produce or sale products ator below cost. The differences between central governmentsubsidy and local government subsidy are the subject and objectof subsidy. Central government subsidy that comes from treasuryis available nationwide. Local government subsidy just adaptslocal enterprises and the subsidy comes from province treasury.In order to elaborate subsidies, we compare SG subsidy withcoal-bed gas subsidy. The central government's subsidy of SG istwo times higher than that of CBM (shown in Fig. 3).

The CBM has been commercialized for a relatively long time,and the local government provides more preferential subsidy. Forexample, Finance Ministry of Shanxi province will offer 0.05 CNY/m3 (0.8 cents/m3) for SG enterprise exploring in this province.Then, CBM's comprehensive subsidy can reach 0.33 CNY/m3

(5.03 cents/m3) after being added other tax breaks.Furthermore, Chinese government has many preferential poli-

cies for SG enterprises. For example, diversied investment isencouraged and the joint venture company is allowed to holdmineral rights. Moreover, the enterprises will enjoy privilegedaccess to land and permits. At the same time, the investment ofdemonstration projects will be simplied. In addition, the devel-opment and utilization of SG may obtain some other scal policiesand taxation policies, such as accelerated depreciation.

4.3. Competition situation

Industrial competitors, potential competitors and quittingcompetitors can reect the competitive conditions in SG industrycomprehensively at the same time. Industrial competitors reectcompetitive condition statically at the present stage. Potentialcompetitors reect the competitive condition dynamically. Whenthe industry shows good momentum of development, there aremore potential competitors. As a result, the increasing number ofpotential competitors may cause intense competition. If thenumber of quitting competitor increases, the market may havevery bearish expectations. According to this principle, we cananalyze the competitive situation of SG as follows:

(1) Market competition situation in 2011SG exploration blocks have been selected for the rst biddingwith invited companies to attend the bidding in 2011. Six

companies were invited to bid for four blocks (Nanchuanblock, Xiangxiu Mountain block, Suiyang block and Fenggangblock) in two provinces, and these six companies are: CNPC,CPDC, CNOOC (China National Offshore Oil Corporation), Oiladministration bureau of Yanchang, CUCMC (China UnitedCoal-bed Methane Corporation) and HNCMC (Henan CoalbedMethane Corporation). Each company can bid for no morethan 2 blocks.Though all these invited companies wish to get the superiorresources, Suiyang and Fenggang blocks in Guizhou provincereceived no more than three biddings. According to biddingrules stipulate, if the number of bidder fails to exceed threebiddings, the tendering and bidding will be invalid.In conclusion, the winning enterprises were both state-ownedenterprises which have advantages of funds, equipment,personnel, and fracturing technologies. The CPDC got oneblock at the price of 591.1 million CNY (97.2 million US dollars)to drill eleven wells. The HNCMC got another block at the priceof 247.6 million CNY (40.7 million US dollars) to drill 10 wells.

(2) Market competition situation in 2012Unlike the rst bidding, the second bidding selected an openbidding mode: domestic enterprises or sino-foreign joint enter-prises, which have oil and gas mineral exploration qualication,and have registered within the territory of the People's Republicof China with the registered capital of 300 million CNY (48.5million US dollars) or more, can bid for SG exploration rights.Each company can tender no more than 2 blocks.This bidding has a total of 20 block distributions in eight regions:Chongqing, Guizhou, Hubei, Hunan, Jiangxi, Zhejiang, Anhui, andHenan. A total of 83 enterprises participated in the bidding andsubmitted 152 copies of the tenders. Higher than the rstbidding, there are 7.6 companies on average competing for oneblock, with state-owned enterprises still playing the leading role,and private enterprises accounting for about one-third (shown inFigs. 4 and 5). What is different with the former bidding is thatall blocks have winning enterprises, in which there are twoprivate enterprises, accounting for 10%.

(3) Market competition situation in futureIn order to reduce winning enterprises' risk, Chinese governmentmay implement the Compensated Transfer Mechanism in thethird SG bidding. It means that the government is trying to helpthe enterprises to do some prophase work for reducing risk andthe bid-winning enterprises should pay the government moneyfor exploration cost. In details, the government will do thepreliminary seismic and drilling exploration to measure the

Fig. 4. SG bidding condition comparison [http://www.mlr.gov.cn/xwdt/jrxw/201107/t20110719_904591.htm]; [http://www.gov.cn/gzdt/2013-01/21/content_2316805.htm].Source: shale gas exploration right bidding results for the rst time and the secondtime.

Fig. 3. Subsidies policy comparison chart of shale gas and coal bed gas.Source: China Treasury and Energy Bureau in document [2012]847.

W. Yunna et al. / Renewable and Sustainable Energy Reviews 45 (2015) 235243240

block actual data rstly. The winning enterprises will repay thegovernment for the cost. Moreover, the government will removethe restriction that enterprises have to own the minimumregistered capital of 300 million CNY (48.5 million US dollars).

5. Simulation results analysis

5.1. Simulation software

The SD model explained in the previous section has been usedto simulate SG market competition using the software of VensimPLE for Windows Version 5.7a for validation of results on SGindustry competition.

5.2. Result analysis

5.2.1. Technology scenarios

(1) Inuence of technology on gas productionTechnological level depends on the investment of research anddevelopment. This paper sets up three research scenarios: highinvestment, standard investment and low investment. Thestandard investment mode is to input more capital in the earlystage. When the technological level reaches a certain level (notmature), the investment begins to reduce. At last, the researchinput keeps stabilized in relatively low level; high investment isto keep high investment. Then the investment is reduced untilthe technology is mature; low investment is to maintain a lowinvestment. It deserves to be noted that Chinese energy marketis inuenced hugely by energy policy since the immature energymarket. For seizing market share and making high prot,investment boom may appear due to a new energy policy. Sothis study assumes the standard input as Z type. It has high inputin rst stage, then slowly decline and keep steady in a range. Thisassumption is more objective.R&D input is divided into two parts: the enterprises' researchinvestment and government's research investment, the formerfocuses on techniques improvement and cost reduction, and thelatter focuses on basic theory and technology research. They havethe similar characteristics and both improve relative technologyof SG. The investment pattern is shown in Fig. 6.R&D investment can increase the technological level, which canimprove the gas production shown in Fig. 7. But the technologycannot have an immediate effect on the gas production, because itneeds time to apply the technology into practice. Gas production

will increase to stable phase quickly under the high investmentmode; the gas production will increase to stable phase slowlyunder the low investment mode. The standard mode changesbetween high and low investment.

(2) Inuence of technology on costTechnological level will improve with R&D investment increas-ing, and the cost will be affected by the technology improving.Cost will come down and reach a steady level from the year 2018very quickly under the high input mode; it will show similartrend under the low input mode, but slower in rate. The standardmode is only a continuum between them. The various inuencedegrees of R&D and technology on cost are shown in Fig. 8.

5.2.2. Cost scenariosTo analyze the inuence of cost on competition, this study

will discuss the trends of potential competitors from three costscenarios by xing other parameters: high cost, standard cost, andlow cost. The high cost is 10% higher than the standard cost, andthe low cost is 10% lower than the standard cost. Seen from Fig. 9,the peak of NPC will appear in 2018 when the cost is dropped by20%; in contrast, the peak of NPC will appear in 2020 if the cost israised by 20%; the standard is a continuum between them. At thesame time, we notice that the cost affects the competition slightlyfrom 2013 to 2016. The reason is that investment risk is still hugeat the early stage when the cost does not decrease greatly. It has alittle inuence on enterprises' behaviors. On the contrary, thecompetition will keep stable under different cost scenarios,because the gas production decides the prot of enterprises when

Fig. 6. Total research investment.

Fig. 7. Gas production trends.

Fig. 5. Winning enterprises condition comparison [http://www.mlr.gov.cn/xwdt/jrxw/201107/t20110719_904591.htm]; [http://www.gov.cn/gzdt/2013-01/21/content_2316805.htm].Source: shale gas exploration right bidding results for the rst time and the secondtime.

W. Yunna et al. / Renewable and Sustainable Energy Reviews 45 (2015) 235243 241

the cost stably shows little reduction space at the later stage,after 2021.

5.2.3. Subsidy scenariosFor analyzing the inuence of government subsidy policy on

competition, we also assume three scenarios of subsidy with xingother parameters: high subsidy (0.8 CNY/m3; 12.95 cents/m3),standard subsidy (0.4 CNY/m3; 6.47 cents/m3), low subsidy(0.2 CNY/m3; 3.22 cents/m3), which is set on relevant policies ofcoal gas and other energy. Then the result of SD model can obtainthe results, shown in Fig. 10.

The results show that different subsidy policies have slighteffect on competition from 2013 to 2017, because subsidy is relatedto gas production according to Chinese subsidy policy. The subsidypolicy cannot bring more prot to enterprises as the gas produc-tion is limited at the early stage. Therefore, the subsidy at the earlystage has little effect on competition. However, when the gasproduction increases, subsidy will have huge inuence on protand will cause erce competition. Furthermore, if the subsidy ishigher, the peak of competition will come earlier. At the laterstage, the subsidy will become relatively less than the prot.Therefore, the impact of subsidy will be weakened again.

5.2.4. Combined inuence of technology, cost and policy on marketcompetition

In this paper, the indicator of NPC is used to study the marketcompetiveness of SG industry by reecting the number of poten-tial competitors. We can nd that the number of competitors willarrive at its peak from 2019 to 2020 and fall back slightly intostable state (shown as Fig. 11). Why does this kind of phenomenon

appear? The strong correlation between NPC and GPR exists.When the GPR reaches the stable stage, NPC will reach the peakvalue. The reason is that the increase of GPR will attract moreenterprises to enter into SG industry. When the GPR is stable, thenumber of enterprises will enter into a cooling-off period. Thenthe NPC will keep stable when the cost and production becomesteady. The number of potential competitors will peak earlierunder high input scenario, because the high input of researchinvestment decreases the uncertainty of technology and cost.

6. Conclusions and suggestions

The SD model has been built to study the trend of Chinese SGmarket competition under different scenarios. According to theresults, we can come to the following conclusions:

(1) Technology becomes the most critical factor inuencing theformation of Chinese SG competitive market. According to theanalysis above, shale gas mining technology inuences notonly extraction cost but also gas production. Furthermore,extraction cost inuencing the amount of prot and gasproduction inuencing the value of the subsidies determinethe living space of enterprises in Chinese SG industry.

(2) The absolute prot remains an overarching concern for allcompetitors. Low extraction cost and high subsidy all willbring an earlier peak of the competitors. The peak will appearin 2018 brought by the former, and in 2019 by the latter. Tworeasons will account for that. Low cost means low barriers toentry including technological barriers and capital barrier andhigh gas production, which stimulates the enthusiasm of

Fig. 10. Inuence of various subsidies on potential competitors.

Fig. 11. Trends for potential competitors.

Fig. 8. Exploitation cost trends.

Fig. 9. Inuence of various cost on potential competitors.

W. Yunna et al. / Renewable and Sustainable Energy Reviews 45 (2015) 235243242

competitors. High subsidy means the high value of the sub-sidies. Thus, more competitors are attracted to the SG industry.

(3) The peak of the competitors will appear at 2019 and 2020according to the combined inuence. When technology, cost andpolicy are all taken into consideration, the peak appears at 2019under high input and at 2020 under standard and low input.

Based on the above conclusions, we present the followingsuggestions:

(1) The number of potential competitors will arrive at the peak in2019. Thus, Chinese government should pay more attention to thenorms and provisions of entry and exit SG industry and a perfectmarket system should be established before 2019, which willprovide basic prerequisite for SG market competition.

(2) Lower mining costs will contribute to the earlier coming ofcompetitive situation. From 2013 to 2018, the high-inputstrategy should be applied by Chinese government urgently.That is to say, a high investment R & D is needed currently,which will bring technological upgrading and lowermining costs.

(3) From 2013 to 2016, the competitors need input high R&D toovercome technical barriers. Due to the high investment, yieldequilibrium point will appear in 2019, which means achieving themaximum subsidy prots and keeping stronger competitiveness.

(4) The competitors need to reasonably apply competition strate-gies. Before 2019, technological barriers are high and thecompetition is weak. After 2020 the technologies will bemature and the competition will be erce. The competitorsshould take their condition into consideration, including fundsand technology, when they prepare to enter or quit.

Acknowledgments

The whole research process is funded by National NaturalScience Foundation of China (71271085), Social Science Fund ofChina (12JGB044) and the Fundamental Research Funds for theCentral Universities of China.

References

[1] Xingkun Qian, Jianshan Xu, Donglong Lin. The oil and gas industry develop-ment report. Beijing: Economic and technical institute of China petroleumgroup; 2013 (in Chinese).

[2] Dittrick P. ETA-ART issues update of world assessment of shale oil, shale gas.Oil Gas J 2013;111(7):46.

[3] Snow N. BP. Shale gas, tight oil to reshape global markets by 2030. Oil Gas J2013;111(2B):25.

[4] Melikoglu M. Shale gas: analysis of its role in the global energy market. RenewSustain Energy Rev 2014;37:4608.

[5] Stevens P. The Shale Gas Revolution: Hype and Reality. London: ChathamHouse; 2010.

[6] Spence D. The shale gas revolution continues. Power 2013;157(2):60.[7] Chang YH, Liu XJ, Christie P. Emerging shale gas revolution in china. Environ

Sci Technol 2012;46(22):122812.[8] Chen SB, Zhu YM, Wang HY, Liu HL, Wei W, Fang JH. Shale gas reservoir

characterisation: a typical case in the southern Sichuan Basin of China. Energy2011;36(11):660916.

[9] Ding WL, Zhu DW, Cai JJ, Gong ML, Chen FY. Analysis of the developmentalcharacteristics and major regulating factors of fractures in marinecontinentaltransitional shale-gas reservoirs: A case study of the CarboniferousPermianstrata in the southeastern Ordos Basin, central China. Mar Pet Geol2013;45:12133.

[10] Lin L, Zhang JC, Li YX, Jiang S, Tang X, Jiang SL, et al. The potential of China'slacustrine shale gas resources. Energy Explor Exploit 2013;31(2):31735.

[11] Wang Q, Chen X, Jhac AN, Rogersd H. Natural gas from shale formationtheevolution, evidences and challenges of shale gas revolution in United States.Renew Sustain Energy Rev 2014;30:128.

[12] Xingang Z, Jiaoli K, Bei L. Focus on the development of shale gas in Chinabased on SWOT analysis. Renew Sustain Energy Rev 2013;21:60313.

[13] Graber R, Retson T. Evaluating the shale gas challenge. Nucl Eng Int 2013;58(705):137.

[14] Guo JC, Zhao ZH. China vigorously promoting shale gas exploration, develop-ment. Oil Gas J 2012;110(3):60.

[15] Hu DS, Xu SQ. Opportunity, challenges and policy choices for China on thedevelopment of shale gas. Energy Policy 2013;60:216.

[16] Jenner S, Lamadrid AJ. Shale gas vs. coal: policy implications from environ-mental impact comparisons of shale gas, conventional gas, and coal on air,water, and land in the United States. Energy Policy 2013;53:44253.

[17] Liu DH, Zha ZG. Shale gas exploitation in China: resource status, technicalbottleneck and solving strategies. In: Proceedings of the third internationalsymposium-industrial engineering and management; 2012. p. 1823. (inChinese).

[18] Schafft KA, Borlu Y, Glenna L. The relationship between marcellus shale gasdevelopment in Pennsylvania and local perceptions of risk and opportunity.Rural Sociol 2013;78(2):14366.

[19] Tollefson J. China slow to tap shale-gas bonanza. Nature 2013;494(7437):294.[20] Forrester JW. Industrial dynamics: a major breakthrough for decision makers.

Harv Bus Rev 1958;36(4):3766.[22] Bodger PS, May DG. A system dynamics energy model of New Zealand.

Technol Forecast Soc Change 1992;41(1):97106.[23] Chowdhury S, Sahu KC. A system dynamics model for the Indian oil and gas

exploration/exploitation industry. Technol Forecast Soc Change 1992;42(1):6383.

[24] Chyong Chi K, Nuttall WJ, Reiner DM. Dynamics of the UK natural gasindustry: system dynamics modelling and long-term energy policy analysis.Technol Forecast Soc Change 2009;76(3):33957.

[25] Feng YY, Chen SQ, Zhang LX. System dynamics modeling for urban energyconsumption and CO2 emissions: a case study of Beijing, China. Ecol Model2013;252:4452.

[26] Naill RF. Managing the discovery life cycle of a nite resource: a case study ofU.S. natural gas. Massachusetts Institute of Technology; 1972.

[27] Movilla S, Miguel LJ, Blzquez LF. A system dynamics approach for thephotovoltaic energy market in Spain. Energy Policy 2013;60:14254.

[28] Ansari N, Sei A. A system dynamics model for analyzing energy consumptionand CO2 emission in Iranian cement industry under various production andexport scenarios. Energy Policy 2013;58:7589.

[29] Kiani B, Ali Pourfakhraei M. A system dynamic model for production andconsumption policy in Iran oil and gas sector. Energy Policy 2010;38(12):776474.

[30] Casey R, Tyli J. Mobile voice diffusion and service competition: a systemdynamic analysis of regulatory policy. Telecommun Policy 2012;36(3):16274.

[31] Sovacool BK. Cornucopia or curse? Reviewing the costs and benets of shalegas hydraulic fracturing (fracking) Renew Sustain Energy Rev 2014;37:24964.

[32] CNFOL. Shale gas tycoon in the United States want to enter Chinese markethttp://www.cnstock.com/08chanye/roll/201301/2438635.htm (in Chinese).

W. Yunna et al. / Renewable and Sustainable Energy Reviews 45 (2015) 235243 243

A system dynamics analysis of technology, cost and policy that affect the market competition of shale gas in ChinaIntroductionSystem dynamics methodSystem dynamics model of SG industryStructure analysis of competition systemModel assumptionVariables and parameters of SD model

Case study and data collectionExploitation costCost estimation of drilling processCost estimation of well completion processCost estimation of well cementation processCost estimation of logging process

Government policyCompetition situation

Simulation results analysisSimulation softwareResult analysisTechnology scenariosCost scenariosSubsidy scenariosCombined influence of technology, cost and policy on market competition

Conclusions and suggestionsAcknowledgmentsReferences