2.7.1.1 Technological change in no-climate policy (reference) scenarios

The importance of technological change for future GHG emission levels and hence the magnitude of possible climate change has been recognized ever since the earliest literature reviews (Ausubel and Nordhaus, 1983). Subsequent important literature assessments (e.g. Alcamo et al., 1995; Nakicenovic et al., 1998b; Edmonds et al., 1997; SRES, 2000) have examined the impact of alternative technology assumptions on future levels of GHG emissions. For instance, the SRES (2000) report concluded technology to be of similar importance for future GHG emissions as population and economic growth combined. A conceptual simple illustration of the importance of technology is provided by comparing individual GHG emission scenarios that share comparable assumptions on population and economic growth, such as in the Low Emitting Energy Supply Systems (LESS) scenarios developed for the IPCC SAR (1996) or within the IPCC SRES (2000) A1 scenario family, where for a comparable level of energy service demand, the (no-climate-policy) scenarios span a range of between 1038 (A1T) and 2128 (A1FI) GtC cumulative (1990-2100) emissions, reflecting different assumptions on availability and development of low- versus high-emission technologies. Yet another way of illustrating the importance of technology assumptions in baseline scenarios is to compare given scenarios with a hypothetical baseline in which no technological change is assumed to occur at all. For instance, GTSP (2001) and Edmonds et al. (1997, see also Figure 3.32 in Chapter 3) illustrate the effect of changing reference case technology assumptions on CO₂ emissions and concentrations based on the IPCC IS92a scenario by holding technology at 1990 levels to reveal the degree to which advances in technology are already embedded in the non-climate-policy reference case, a conclusion also confirmed by Gerlagh and Zwaan, 2004. As in the other scenario studies reviewed, the degree to which technological change assumptions are reflected in the scenario baseline by far dominates future projected emission levels. The importance of technology is further magnified when climate policies are considered. See for example, the stabilization scenarios reviewed in IPCC TAR (2001) and also Figure 2.1 below.

Figure 2.1: Emission impacts of exploring the full spectrum of technological uncertainty in a given scenario without climate policies. Relative frequency (percent) of 130,000 scenarios of full technological uncertainty regrouped into 520 sets of technology dynamics with their corresponding carbon emissions by 2100. Also shown is a subset of 13,000 scenarios grouped into 53 sets of technology dynamics that are all ‘optimal’ in the sense of satisfying a cost minimization criterion in the objective function. See text for further discussion. 1 Gt C = 3.7 Gt CO₂

Source: Adapted from Gritsevskyi and Nakicenovic, 2000.

Perhaps the most exhaustive examination of the influence of technological uncertainty to date is the modelling study reported by Gritsevskyi and Nakicenovic (2000). Their model simulations, consisting of 130,000 scenarios that span a carbon emission range of 6 to 33 GtC by 2100 (Figure 2.1), provided a systematic exploration of contingent uncertainties of long-term technological change spanning a comparable range of future emissions as almost the entirety of the no-climate policy emissions scenario literature (see Chapter 3 for an update of the scenario literature). The study also identified some 13,000 scenarios (out of an entire scenario ensemble of 130,000) regrouped into a set of 53 technology dynamics that are all ‘optimal’ in the sense that they satisfy the same cost minimum in the objective function, but with a bimodal distribution in terms of emissions outcomes. In other words, considering full endogenous technological uncertainty produces a pattern of ‘technological lock-in’ into alternatively low or high emissions futures that are equal in terms of their energy systems costs. This finding is consistent with the extensive literature on technological ‘path dependency’ and ‘lock-in phenomena’ (e.g. Arthur, 1989) as also increasingly reflected in the scenario literature (e.g. Nakicenovic et al., 1998b and the literature review in Chapter 3). This casts doubts on the plausibility of central tendency technology and emissions scenarios. It also shows that the variation in baseline cases could generate a distribution of minimum costs of the global energy system where low-emission baseline scenarios could be as cheap as their high-emission counterparts.

The results also illustrate the value of technology policy as a hedging strategy aiming at lowering future carbon emissions, even in the absence of directed climate policies, as the costs of reducing emissions even further from a given baseline are ceteris paribus proportionally lower with lower baseline emissions.