In this issue

  Results

Down the learning Curve with Emerging Technologies

Briefing Note on the Kyoto Protocol

Japan needs Nuclear Power to Reduce CO2 emissions

Markal ‘‘most widely used’’ model, but...



Click on image!

Figure 2. Global electricity generation by technology in the base scenario: case 1, a local optimum with business-as-usual.



Click on image!

Figure 3. Global electricity generation by technology in the base scenario: case 2, a global optimum with a more diverse system.



Click on image!

Figure 4. Higher early investment costs in case 2 may lock out emerging technologies.



Click on image!

Figure 5. Global electricity generation by technology in the limited carbon dioxide emissions scenario.



Click on image!

Figure 6. Annual CO2 emissions from the global electricity system in the GENIE model.


 

Results were produced first for a base scenario, for which two different solutions can be obtained. The first solution, a local optimum, appears in Figure 2. It can be described as a business-as-usual development of the global electricity system, with total system costs amounting to $9,117 billion. In this solution, conventional fossil technologies are phased out and initially replaced by CCGT and hydro-power. Later, possibly due to increased gas prices, CCGT is replaced by advanced coal power, which eventually becomes the dominant technology of the system. Carbon dioxide emissions from this system almost double by the middle of the next century, as shown in Figure 6.

A completely different solution to the same scenario appears in Figure 3. This is the true optimal (least cost) solution which has a total system cost of $9,106 billion, marginally lower than the first case. In case 2, fuel cells swiftly gain market share and eventually become the largest source of electricity. Photovoltaic cells (PV) contribute substantially to global electricity production, and non-intermittent PV-H2 also enters the system. Total carbon dioxide emissions increase by a maximum of 30 percent, but are later reduced below 1995 levels.

The lower costs, lower emissions and increased technological diversity of case 2 suggest that this path can be viewed as a no-regrets policy, making it the preferred choice. But the choice must be made early: in case 1, there are no investments in PV or fuel cells. In case 2, these technologies grow at the maximum rate from the first time period on. During the first decades, however, these investments are not profitable, but they are necessary to ensure future (greater) profitability. This situation is illustrated in Figure 4, which shows annual investment cost profiles for the two solutions.

This figure emphasizes the risk of technology lock-in. Case 2 requires approximately 30 percent more investment capital than case 1 in the year 2025. If capital is a scarce resource in the future, a fairly safe assumption, there is danger that capacity will be built up with established technologies as in case 1, the business-as-usual future. There will then be no opportunity to gather cost-reducing experience with emerging technologies because they will be effectively locked out by established technologies.

“However”, says Mattsson, “implicit in the model representation is the assumption that large grid-connected electricity systems will bear the costs of introducing the emerging technologies. In practice, nursing and bridging markets may provide a natural growing ground for the emerging technologies. The burden of technology development on the grid-connected systems may then be eased and lock-in prevented”.

In a second “greenhouse” scenario, a limit was placed on accumulated emissions of carbon dioxide equivalent to 50 years of emissions as the current level, with the resulting emission trajectory shown in Figure 6. A local optimum is shown in Figure 5. The value of the emerging technologies PV and fuel cells is much higher in the limited carbon dioxide emissions scenario, and consequently both technologies are developed as quickly as possible in both cases. Total system costs are $9,232 billion. Thus, the demanding limits on carbon dioxide emissions raise the cost only slightly more than 1 percent.

To obtain a local optimum which is different from the global least-cost solution, Mattsson introduces extra constraints that make the global optimum infeasible and force the solver to find another optimum. For example, case 2 in the base scenario is the true global optimum. To obtain case 1, PV and fuel cell capacity were temporarily constrained (to less than 20 GW in 2025) but not explicitly forbidden. The constrained solution was not to invest in these technologies at all. Since the capacity constraint of 20 GW was inactive, this was a local optimum to the original problem. This can occur in a nonlinear problem with several local optima, whereas in a linear problem any such additional constraint would automatically be binding.

“The GENIE results illustrate that plausible locally optimal solutions can show deviating paths that lead to drastically different future energy systems”, says Mattsson. “Timely support of emerging technologies is probably necessary to avoid lock-in of established technologies and build a diverse, flexible energy system”.

Mattsson and Wene plan to study future electricity options for Sweden by soft-linking GENIE with MARKAL. To facilitate this linking, GENIE was designed to be as compatible with MARKAL as possible.

References

N. Mattsson. Internalizing technological development in energy systems models. Thesis for the degree of licentiate in engineering. Chalmers University of Technology. November 1997. N. Mattsson and C.-O. Wene. Assessing new energy technologies using an energy system model with endogenized experience curves. International Journal of Energy Research, Vol. 21, 385-393 (1997).
R.H. Williams and G. Terzian. A benefit/cost analysis of accelerated development of photovoltaic technology. PU/CEES report no. 281, Princeton University, New Jersey, USA.

Go to Briefing Note on the Kyoto Protocol