Planning task D: Impact estimation
Example: Local energy planning in Hylkje

Four planning alternatives have been identified as relevant in Hylkje area. The planners decided on five criteria to use when comparing among these alternatives: operating and investment costs, CO2 emissions, NOx emissions and heat dump.

The estimation of the impacts alternatives will have in each of these five criteria has been a challenging task because of the large amount of data that had to be processed. For this case study a computer simulation/optimization model of the energy system has been used. This model is called eTransport and has been developed at SINTEF Energy Research in Trondheim, Norway. eTransport is a user-oriented flexible and easy-to-use tool to support decision making. Energy systems can be easily modelled within the tool, and a variety of analyses can be performed. The eTransport model is one of the tools included in the category 'Energy system models' above. By following the links you will be led to a more detailed description of the use of this model.

All four alternatives have been simulated using eTransport. The following simplifications and assumptions have been made when using this model.

A 122 bus network was used for the electricity grid, with hourly electricity load specified in 55 of them. DC load flow equations were used to calculate the load flow and corresponding losses in the impact model. Potential district heating networks were represented with either 14 or 16 heat demand points, all of them with hourly demand data for the 8 load days. Note that while the electricity load can only be met by electricity, any connected energy carrier can meet the heat load. In this case that is electricity or district heating.

The eTransport impact model finds the minimum cost solution for meeting both electricity and heat loads for each of the days considered.

The main uncertainty considered in the analysis is the price of electricity. The electricity price is very important for the total cost of meeting the load, since there can be substantial exchange of electricity from the area, both imports and exports. Three scenarios (high, base, low)are used for hourly prices of electricity as shown in the following figure (the prices are given here in NOK - Norwegian Kroner).

In addition to the price uncertainty, it has also been assumed that the marginal change in global CO2 emissions from exchange of electricity was uncertain. This factor affects the total CO2 emissions from different planning alternatives. The marginal CO2 factors for electricity exchange were set to 400, 500 and 600 g/kWh respectively, for the low, medium and high price scenarios, assuming that more efficient technologies are used in the low price scenario. Subjective probabilities were assigned to the scenarios, using 0.25 for the high and low scenarios and 0.5 for the medium price scenario. These probabilities were used when calculating the expected utilities.

Other prices, such as the price for gas supply to CHP plants and gas boilers, and the price paid for heating at the industrial site were assumed constant in the analysis.

The impacts for all four alternatives in the five criteria and three scenarios considered are summarized in the table below.

The next step is that the planner will have to choose one of the alternatives based on the calculated impacts. We can see from the table that alternative 1 has higher operating cost and CO2 emissions than the three other alternatives. On the other hand, the investment cost and the local emissions of NOx and heat are lower in scenario 1. The differences between the last three scenarios are smaller, but still significant, especially for NOx emissions and heat dump. There are also differences in the level of uncertainty for the attributes in the four alternatives, as can be seen when studying the results from the three price scenarios.The planners could of course base their decision on direct assessment of the information in the table above. However, even with the simple example presented here it becomes difficult to judge the planning trade-offs and risks. The next step in this example shows how the decision can be optimized with respect to all criteria relevant in planning.

Published August 3, 2012

Contact: Gerd Kjølle, SINTEF Energy Research