Geographical islands can play a crucial role to become a forerunner towards more sustainable and distributed energy systems, since they are isolated and are usually dependent on fossil fuels imported from the mainland. Here, our results present the optimal design of multi-energy systems (MESs) on such geographical islands for a case study in Eigerøy (Norway). A mixed integer linear program is formulated and solved to minimize costs and greenhouse gas (GHG) emissions while meeting the energy demands of given end-users.

Figure 1 shows a schematic of a multi-energy system that comprises hydrogen, natural gas, syngas, and (renewable) electricity as energy sources and carriers and uses a wide set of energy technologies considering grid-connected configurations. The figures and content herein (on this webpage) were published in: Applied Energy, 347, Terlouw, T., Gabrielli, P., AlSkaif, T., Bauer, C., McKenna, R. & Mazzotti, M. (2023). Optimal economic and environmental design of multi-energy systems, 121374, Copyright Elsevier (2023) [1] licensed under CC BY 4.0[2].

 

Figure 1 MES installed, here on Eigerøy. Figure is obtained from Terlouw et al.¹ 

Scenarios

 

For the comparison between the current energy system (business-as-usual, i.e. BAU) and potential design scenarios, the following optimization scenarios are considered (see x-axis in Figure 2):

• (1) Scenario “BAU”: represents the current energy system of Eigerøy in 2021, mainly consuming fossil fuels for heat and low-carbon electricity from the Norwegian grid.
• (2) Scenario “Cost-Min-Res”: is a minimum-cost optimization, which excludes environmental considerations in the objective function and excludes high-temperature heat requirements and personal transport (but includes power requirements of the geographical island).
• (3) Scenario “Cost-Min-Res-M”: is a minimum-cost optimization, which excludes environmental considerations in the objective function and excludes high-temperature heat requirements.
• (4) Scenario “Cost-Min”: is a minimum-cost optimization, which excludes environmental considerations in the objective function.
• (5) Scenario “Cost-GHG90”: is a minimum-cost optimization, with a constraint of life cycle GHG emissions to reach a reduction of 90% compared to a cost optimization (which can be obtained on the Pareto front).
• (6) Scenario “GHG-Min”: is an optimization of life cycle GHG emissions, which excludes cost considerations in the objective function.

 

Figure 2. Main results for costs and life cycle GHG emissions. Figure is obtained from Terlouw et al.¹

Results

 

Figure 2 illustrates the major cost and emissions contributions to the overall expenses and life cycle GHG emissions across all considered scenarios in Eigerøy (see previous paragraph), with each segment of the stacked bars represented in different colors. The scenarios are displayed on the x-axis, annual costs on the primary y-axis, and life cycle GHG emissions on the secondary y-axis. For each scenario, the left bar shows the cost contributions, while the right bar depicts the life cycle GHG emissions. The values above the bars indicate the investment contributions to the overall costs (left bar) and the contributions of embodied emissions to the overall life cycle GHG emissions (right bar). Please note that the second and third scenarios exclude high-temperature heat; the second scenario (Cost-Min-Res) also excludes personal transport. These scenarios are highlighted with a light grey shaded area.

 

This latter figure highlights the following findings. First, the optimal design relies heavily on integrating different energy sectors, shown by the shift from scenario BAU to Cost-Min in Figure 2. Including the industrial and mobility sectors more than doubles the costs and life cycle GHG emissions, requiring additional energy technologies, such as natural gas boilers, advanced CHP units, and hydrogen systems.

 

Second, Figure 2 reveals that the construction phase can significantly impact total environmental burdens, contributing up to 60% of GHG emissions in low-carbon MES designs. However, this impact could reach 80%, e.g. for human toxicity, metals & minerals, and ozone depletion, highlighting the need to consider embodied (life cycle) emissions in MES analyses.

 

Figure 3. Pareto front for the case study in Eigerøy. Figure is obtained from Terlouw et al.¹

References

 

[1] Terlouw, T., Gabrielli, P., AlSkaif, T., Bauer, C., McKenna, R., & Mazzotti, M. (2023). Optimal economic and environmental design of multi-energy systems. Applied Energy, 347, 121374, 10.1016/j.apenergy.2023.121374.

[2] https://creativecommons.org/licenses/by/4.0/.