A Multi-Stage Integrated Planning Framework for Electricity and Natural Gas under High Renewable Energy Penetration

Abstract

Increasing use of energy affects the environment and reduces the availability of global resources. To effectively manage the available energy resources of a country, energy planning tools have been used as key decision-making instruments to support investment decisions and to assess long-term implications of the future options for the system, economy, environment, and society.Power system planning is about ensuring that investments in new generation, transmission and distribution facilities are sufficient to meet the future demand requirements in an economical and reliable manner. However, traditional power system planning must evolve to deal with many challenges coming from the deregulation of the electrical power industry, global warming and the lack of complete and accurate information about the future.The deregulation of the electrical power industry has introduced competition in the generation and retail sectors and has created electricity markets around the world, which in turn introduces uncertainties about spot prices and generation costs. These uncertainties are intensified by the difficulty in making long-term forecasts and predicting the unfolding of uncertainties over time. Global warming has encouraged the development of renewable energy sources and, to some extent, provided incentive to accelerate gas-to-power projects worldwide. Although many of these renewable energy sources are clean, they have high capital cost requirements and some have highly uncertain production for example, in the case of solar and wind generation, dependent on weather conditions. Natural gas is more competitive among the traditional fossil fuels in a carbon-constrained economy, leading to the proliferation of gas-fired power generation (GPG) in the power sector. The increasing share of gas-fired power generation results in a stronger interdependency between electricity and natural gas systems that might affect the natural gas supply adequacy and lead to electricity disruption. Thus, power system planning must be carried out jointly with gas system planning, taking into account the link between these two systems and the economic and physical aspects of gas. In addition, a combined decision-making model is fundamental in cases where the same energy system operator jointly operates both the electricity and gas markets and systems as is the case of the Australian Energy Market Operator (AEMO).This research aims to take into consideration these new challenges that have been introduced into traditional power system planning. A comprehensive multi-stage (dynamic) model is developed to simultaneously co-optimize power and gas systems while dealing with variable renewable energy resources (VRE). The model employs stochastic programming to address the uncertainty of load growth, renewable technology development, gas prices and renewable energy production. The proposed multi-stage model provides a more faithful representation of real-world decision-making by allowing multiple investment decisions over time. This framework is designed to adapt investment and operation decisions to the unfolding of uncertainties faced by energy and gas systems over time.The impact of high renewable energy penetration is assessed on a realistic case of the State of Queensland, Australia, whereby equivalent electricity and gas networks are designed to capture the link between these systems and to accommodate the staters unique availability of renewable energy resources. A number of cases are simulated for both individual electricity planning and integrated planning of the electricity and gas systems.All problems are cast as (stochastic) linear programming implemented using the General Algebraic Modelling System (GAMS) mathematical programming language, and solved using the IBM/CPLEX solver.The model results demonstrate that performing an integrated planning of electricity and natural gas presents benefits over separate planning of each network among which is a lower overall cost to expand both systems and more gas supply security as the gas constraints are known.Findings of the dynamic approach show that more robust outcomes are obtained over the static and rolling-window static approach. First, as it considers internal stages of the planning horizon as opposed to designing a static system for the end year and presents a global optimization window instead of successively optimizing each sub-problem separately as in the rolling-window framework. Second, allows the expansion plans to be adapted to future unexpected changes, reducing the risk of underestimating investments in new power capacity as in the static approach and overestimating as in the rolling-window framework.nn