توسعه یک مدل پیش بینی نفوذ در بازار برای خودروهای سلول سوختی هیدروژنی همراه با در نظر گرفتن زیرساخت ها و اثرات کاهش هزینه

In order to cope with climate change, the development and deployment of Hydrogen Fuel Cell Vehicles (HFCVs) is becoming more important. In this study, we developed a forecasting model for HFCVs based on the generalized Bass diffusion model and a simulation model using system dynamics. Through the developed model, we could forecast that the saturation of HFCVs in Korea can be moved up 12 years compared with the US. A sensitivity analysis on external variables such as price reduction rates of HFCVs and number of hydrogen refueling stations is also conducted. The results of this study can give insights on the effects of external variables on the market penetration of HFCVs, and the developed model can also be applied to other studies in analyzing the diffusion effects of HFCVs.

The global warming problem due to climate change has been threatening the sustainable development of all humankind (IPCC, 2007). In order to cope with climate change, many new technologies that can reduce greenhouse gas (GHG) emissions have been developed and deployed (IEA, 2008). The Hydrogen Fuel Cell Vehicle (HFCV) is especially considered as the most promising technology to reduce GHGs emissions in the transportation sector (Barreto et al., 2003, Clark and Rifkin, 2006, Dunn, 2002 and Winebrake, 2002). Therefore, the demand for studies to forecast the market penetration of HFCVs has increased to establish successful commercialization strategies for HFCVs (McDowall and Eames, 2006). The most widely used model to forecast the demand for new products is the Bass diffusion model because it can describe the S-shape penetration curve of new products with meaningful parameters such as the innovation factor and imitation factor (Bass, 1969, Mahajan et al., 1990 and Nigel and Towhidul, 2006). However, the Bass diffusion model cannot consider external variables that can affect the market penetration of new products such as marketing effort and cost reduction, so the generalized Bass diffusion model was proposed (Bass et al., 1994). The greatest difficulty when applying the diffusion model to a real problem is that there is not available data to estimate model parameters because the target of forecasting is a new product that has not yet been introduced to the market. For this reason, most of the studies related to the market penetration forecasting of HFCVs have used a discrete choice model based on survey results, and the generalized Bass diffusion model using historical time series data has not yet been applied to HFCVs (Green, 2001; Paulus, 2008). Gustavo developed a forecasting model for HFCVs using a logistic diffusion model, but the developed model could not analyze the effects of external variables such as infrastructure and cost reduction (Gustavo, 2007). Therefore, the main purpose of this paper is to develop a new forecasting model for the market penetration of HFCVs, based on a generalized Bass diffusion model using alternative products such as the Hybrid Electric Vehicles (HEVs). In addition, we develop a simulation model for the market penetration of HFCVs, using a system dynamics approach, which can analyze system behavior based on the positive and negative feedback effects among system components. Through this study, we can provide insight as to the effects of external variables such as infrastructure and cost reduction on the market penetration of HFCVs, and we can develop a simulation model that can enhance the versatility and practicality of the forecasting model. The results of this study can promote research related to the market penetration forecasting for HFCVs using the generalized Bass diffusion model, and the developed simulation model can be applied to other studies in analyzing the effect of HFCVs diffusion.

In this study, we developed a penetration forecasting model for HFCVs in Korea based on the generalized Bass diffusion model. Parameters of the model were estimated from the time series data, including the sales volume of the HEV in Japan, the price ratio between the Prius and the Corolla, the number of LPG fueling stations in Korea, the number of LPG vehicles in Korea and the number of total light duty vehicles in Korea. The comparison results of the penetration of three countries, Korea, the U.S. and Japan, indicate that market saturation of Korea can be achieved 12 years faster than that of the U.S. It seems that Korea’s more active and risk-taking purchasing patterns for innovative products are reflected in the imitation factor. Because the diffusion performance of the hybrid vehicles in Japan was applied as one of the basic data for the development of the demand forecasting model of HFCVs in this study, the sensitivity analysis was conducted with the estimated innovation factor and imitation factor. The results showed that the innovation factor and the imitation factor have direct influences on the time period in reaching the critical mass and market saturation, respectively. Considering that the decision criterion of the success of a new product in the market is the implementation of critical mass, it was judged that various policies would be required to promote the initial purchase probability of the innovators. When the innovation factor was 0.005 or above, it was shown that the time period for reaching the critical mass did not vary much. It was expected that, when the purchase probability of the innovators is more or less than 0.0037, which was the performance of the hybrid cars in Japan, HFCVs also would be able to reach its critical mass within 7–8 years from market introduction. When the purchase probability of the imitators for HFCVs is 0.3454, which was the performance of the hybrid cars in Japan, the new sales would be the most in 2030 by about 700,000 HFCVs. By 2040, all the new cars sold in the market would be HFCVs, and assuming the life cycle of cars to be 10 years, all the vehicles running on the roads of Korea would be HFCVs by 2050. Next, the results of the sensitive analysis on the influences of the price level and number of hydrogen fueling stations on the market saturation of HFCVs showed that the annual reduction rate of the price level compared with that of the internal combustion engine vehicles and the number of new fueling stations built per year would have major influence on the purchases of the imitators and innovators, respectively. We interpreted this as attributable to the idea that, since HFCVs are superior to conventional vehicles in environmental friendliness, and also noise and vibration levels in the driving system, a higher price level than conventional vehicles would not be an obstacle of purchase for the innovators, but, due to the characteristic of cars being highly dependent upon fueling infrastructure, the inconvenience of refueling would raise an obstacle to the innovators. Therefore, it was reasoned that the provision of the initial infrastructure would be more important than reducing price for the successful market penetration of HFCVs. It was analyzed that to prevent delay in reaching the critical mass caused by insufficient infrastructure, 130–260 new fueling stations should be constructed annually. In addition, to promote the market penetration of HFCVs, it is recommended to implement various incentives including tax reductions and financing for initial investment for the construction of hydrogen fueling stations. The significance of this study can be classified into four aspects; first, a more objective and quantitative demand forecasting model was developed on the basis of historical series data, while most of the previous studies on demand forecasting have been based on survey results, which were based on the subjective opinions of people; second, an analysis technique that could broaden the applicability of demand forecasting models was proposed by constructing a simulation model using system dynamics; third, while the existing system dynamics models implemented a feedback structure based on the intuition of the researchers, this study constructed a system dynamics model based on the proven mathematical form of the generalized Bass model to propose the theoretical basis of the model, and fourth, the empirical study using simulation model shows that the initial purchase by innovators is a key factor for the successful market penetration of HFCVs, and the initial infrastructure can play more important role to promote the initial purchase by innovators rather than reducing the price of HFCVs. However in order to acquire more realistic forecasting results, some assumptions underlying the forecasting model need to be released. First, this study makes the assumption that the penetration pattern of HFCVs will be similar to that of the HEV and various sensitivity analyses were conducted based on the innovation factors and imitation factors of the HEV. Therefore, a more reasonable way to estimate the innovation factor of HFCVs needs to be studied. Studies on consumer preferences regarding green cars can be applied to develop a model that can consider the differences between the penetration of the hybrid car and HFCVs. Second, we also assume that conventional vehicles will be replaced by HFCVs completely. However, there are other options besides green cars such as plug-in HEVs and electric vehicles. Hence, the competition of rival technologies has to be considered in the forecasting model. This study also developed a simulation model based on system dynamics in order to enhance usefulness of the model. Through the simulation model, the change of penetration rate resulted from changing parameter values such as the innovation factor, the imitation factor, and the number of hydrogen fueling stations can be measured instantly, and sensitivity analyses on the parameter values can be conducted easily. However, the strongest point of the system dynamics is that it can consider the feedback effect among variables. For example, the construction of the hydrogen fueling stations will lend positive effects to the purchasing of HFCVs. In addition, the purchasing of HFCVs also can confer positive effects on the construction of hydrogen fueling stations. This positive feedback relation can reinforce the effects of the construction of hydrogen fueling stations; so analyses excluding consideration of the feedback relations can underestimate the effects of these variables. Hence, further study on the feedback effects among variables will be necessary, and a simulation model that can consider the feedback effects among variables also needs to be developed.