Reducing energy system model distortions from unintended storage cycling through variable costs

Area of Science:

• Energy Systems Modeling
• Computational Optimization
• Policy Analysis

Background:

• Energy system models are crucial for policy decisions and technology design.
• Improper model usage can lead to implausible outputs, such as unintended storage cycling (simultaneous charging and discharging).
• Existing methods to correct these issues are often computationally inefficient or ineffective.

Purpose of the Study:

• To investigate the impact of variable cost allocation and solver accuracy on unintended storage cycling in energy system models.
• To propose a method for reducing the magnitude and likelihood of unintended storage cycling.
• To explore the potential of this approach in mitigating other similar modeling artifacts.

Main Methods:

• Conducted 124 simulations varying variable cost allocations and solver accuracy levels.
• Analyzed the relationship between these parameters and the occurrence of unintended storage cycling.
• Developed recommendations for variable cost inputs and identified a minimum threshold for mitigation.

Main Results:

• Determining appropriate variable cost levels is dependent on cost allocation and solver accuracy.
• Loose solver accuracy settings can prevent the removal of unintended storage cycling.
• Recommended variable cost inputs and a minimum threshold were identified to reduce unintended storage cycling.

Conclusions:

• The proposed approach effectively reduces unintended storage cycling by optimizing variable costs and solver accuracy.
• This method offers a computationally efficient alternative to existing techniques.
• The findings suggest the approach can also address other misleading effects like unintended line or sector cycling.