**3. Results**

Resulting from the analysis, the following answers to research questions were proposed:

A1 (Q1): According to the methodology, the analysis is feasible in four countries, that is, Spain, the United Kingdom, Norway, and Turkey. In 2025, there will be 10,761 ZEBs in Spain, 5530 in the United Kingdom, 12,658 in Norway, and 2373 in Turkey. However, in 2030, the number of zero-emission buses in the same countries will increase significantly and will amount to 37,854 for Spain, 25,056 for the United Kingdom, 16,267 for Norway, and 25,372 for Turkey.

A2 (Q2): With this predicted number of clean buses, it seems that only Norway will be able to reach 95% level of ZEBs share in all buses possessed by this country.

A3 (Q3): On the basis of analyses conducted, the majority of EU members will have a 95% share of ZEBs in a fleet consisting of all types of buses after 2050. Detailed data are presented in the Table 5 with the forecast of the ZEBs' share in the market in the EU countries in 2025 and 2030. Figure 2 presents the geographical distribution of the results.

Table 5 provides a forecast of the share of ZEBs in the market in 2025 and 2030 for all European countries for which data were available. In addition, it presents in which year the number of buses in this category will constitute 95% of all buses (*m*). Out of the concern about the quality of the model, the results are marked with different colors. Countries with poor fit parameters are marked in gray, while those in which the quality of the model fits the data very well are shown in green (details described in the Materials and Methods chapter). The presented results are negatively affected by the following factors: short reporting period, data quality, and the issue of technology definitions that are inconsistently interpreted by different countries.


**Table 5.** Development forecast for the electricity, hybrid diesel-electric, plug-in hybrid diesel-electric, and hydrogen and fuel cells bus market [own study based on data retrieved from Eurostat].

**Figure 2.** Geographical distribution of the year when countries reach the 95% of ZEBs in their fleets [own study].

The analysis traced the situation of 22 EU countries (Table 5). The detailed analysis of the data showed that only four countries out of all the countries considered show activity related to the replacement of their bus fleet with electric ones. The reasons for this endeavor can be twofold. Either it results from a high level of environmental awareness of the mentioned countries, such as Norway, or it testifies to the countries' high commitment and efficiency in obtaining EU subsidies. The lack of reliable data, including consistent historical sequences for 2013–2018, in the case of the remaining countries may indicate a low level of their activity in this area.

Table 5 contains columns presenting the market share for 2025 and 2030 calculated on the basis of the Bass model. The percentage values are related to the total market share (*m*). The number of buses in individual countries was estimated based on the average number of all buses in a given country for 2013–2018. Additionally, the year in which market penetration by new bus generations will reach the level of 95% is indicated. The 95% level was chosen arbitrarily and results from the slow growth of the S-shaped Bass curve at the end of a given technology development.

Figure 3 presents a histogram, based on the calculations. It indicates the years when traditional buses should be replaced by buses using electricity, hybrid diesel-electric, plug-in hybrid diesel-electric, and hydrogen or fuel cells.

**Figure 3.** Histogram of years of 95% market adaptation by ZEBs (all EU countries included) [own study].

It should be noted that all European countries were considered in Figure 3 regardless of the quality of the model. The chart shows that the average adaptation to the market should take place around 2077 (average), with a standard deviation of about 28 years. In addition, the above histogram shows compliance with the normal distribution based on the Kolomogorov–Smirnov test, with *p*-value 2.23745 × 10−<sup>7</sup> and statistic 0.473591; however, the Shapiro–Wilk test gives statistic 0.973548 and *p*-value 0.56575.

In the case of countries for which the coefficient of determination R<sup>2</sup> was higher than 0.9 (see Table 3), the histogram is presented in Figure 4.

**Figure 4.** Histogram of years of 95% market adaptation by ZEBs (country with R<sup>2</sup> > 0.9 included) [own study].

The countries were divided into two groups with a higher and lower imitation coefficient. A higher imitation factor *q* > 0.3 means that countries are adopting the new technology relatively quickly. Countries classified in this area include Belgium, Spain, Portugal, Romania, the United Kingdom, Norway, and Turkey, with the average of full adaptation in 2038 and a standard deviation of about 7 years. In contrast, the second group with a lower q usually close to 0.1 constitutes the following countries: Denmark, Germany, France, Croatia, Hungary, Poland, Slovenia, Finland, and Sweden, with the average of around 2095 and a 13-year standard deviation. However, statistics for parameters in these countries give a low level of confidence for calculations of the Bass model variables. The forecast was based on available data. The country in which the forecast indicates a very distant time of market acceptance has considerable uncertainty in estimating this year of adjustment. This is partly owing to the fact that countries have not shown significant activities in this area. It should be noted that, if a given country has already begun investment in a given technology and the process of di ffusion of innovation, then adjustment could take place quite quickly. With the data we have, there is no basis to assess what will happen in a given country if it changes its policy and makes significant investments in ZEBs (for a more detailed comment, see the Materials and Methods section).

The most reliable results were obtained for countries marked in green (Table 5). They have the best parameter estimators and a very good model fit factor. In Spain, the United Kingdom, Norway, and Turkey, the average saturation of the market with zero-emission buses should occur around 2037. The process of technology adaptation calculated from the Bass model is presented in Figure 5. The vertical axis presents market adoption expressed as a percentage and the horizontal axis represents time in years. The cumulated number of buses for selected countries allows for the assessment of innovation di ffusion. For example, on the basis of Figure 4, in 2030 in Norway, the percentage saturation of ZEBs in the total bus transport fleet will reach around 50%.

**Figure 5.** The cumulative Bass curves for Spain, the United Kingdom, Norway, and Turkey [own study].

The development forecast for the entire European Union was made in two versions (Figure 6). The first optimistic variant assumes a high imitation factor. Parameters of the Bass model were calculated on the basis of the average for the best four models, that is, Spain, Norway, the United Kingdom, and Turkey (*p* = 0.000173, *q* = 0.5195). The pessimistic variant was created on the basis of average parameter values for all countries from Table 5 (*p* = 0.001407, *q* = 0.2178). The above approach results from the fact that an attempt to compile data for all EU countries gave a model with unsatisfactory estimators (please see details in Materials and Methods).

It has to be stated that the predicted number of buses could be estimated only for the chosen EU members. This is owing to the lack of a uniform definition of zero-emission vehicles. An additional factor causing calculation di fficulties was errors in Eurostat statistics. In addition, there were gaps in the data collected for individual countries. Thus, the authors could only conduct the correct simulation for four countries: Spain, Great Britain, Norway, and Turkey, as shown in Table 5.

**Figure 6.** The cumulative Bass curves for Spain, the United Kingdom, Norway, and Turkey [own study].
