Introduction

NTU Start Up Grant (2018 - 2019): Linking Waste Heat Recovery and Electrical Power Networks for Energy Efficiency in Eco-Industrial Parks with Heterogeneous Generation Sources

Abstract: Eco-industrial parks (EIPs) have attracted a lot of research attention in recent years. EIPs are industrial parks (such as Singapore’s Jurong Island and South Korea’s Yeosu Industrial Complex) where businesses cooperate with each other and sometimes with the local community to reduce waste and pollution, efficiently share resources and minimize environmental impact while simultaneously increasing business success. A key research area in the context of EIPs is the sharing of energy resources. Combined cycle power plants (CCPPs) are key components which aid in linking the electrical energy and thermal energy streams in EIPs. With growing environmental concerns, future energy systems are likely to be heterogeneous in nature with CCPPs combining with energy storage systems (both electrical and thermal) renewable energy sources and other conventional generators to meet the thermal and electrical energy demands in EIPs. While many industrial plants have plant level waste heat recovery (WHR) networks, there is a lot of scope for improving energy efficiency by operating park level WHR networks. While a few researchers have studied park level WHR networks, the interaction between the electrical generation network and its link to the WHR network via CCPPs has not been investigated thoroughly. This project aims at addressing this gap from various perspectives including but not limited to the following: i) optimal planning of WHR networks with some nodes represented by CCPPs ii) optimization models for scheduling of thermal energy and electrical energy production (multi-energy management systems) considering WHR and electrical distribution networks iii) development of energy trading models using agent-based approaches v) impact of natural gas supply networks on the operation of generators in EIPs iv) frequency regulation using distributed control approaches vi) demand response and load management strategies for efficient energy management vii) pilot scale implementation of control and optimization strategies using OPAL-RT and physical microgrid setup.

MOE AcRF Tier 1 (2019 - 2021): Quantification of Prediction Intervals for Renewable Energy Sources and Scalable Algorithms for Distributed Energy Trading under Uncertainties

Abstract: Singapore is seeking to diversify its energy sources, and involving more cost-competitive suppliers into the national energy market to reduce electricity prices. Due to the proliferation of distributed energy storage and demand response program, the distribution network has become more competitive, where energy consumer can choose from various energy suppliers to purchase the cheapest power, while energy suppliers can also sell their energy to consumers that bid the highest price. This calls for a platform to facilitate the energy trading. A centralized electricity market frequently used in the transmission system is not suitable for this scenario due to the high density of agents within distribution networks. Therefore, we expect to build a peer-to-peer (P2P) energy trading platform for the distribution network. The distributed feature of this P2P energy trading platform is suitable for the distribution network for the following reasons: i) No centralized decision unit is required to communicate and process a tremendous amount of information from the agents, as only distributed communication is required among adjacent agents; ii) energy trading only happens among adjacent agents, which can avoid excessive power losses due to long-distance energy transfer; and iii) the distributed scheme is plug-and-play for heterogeneous agents. In practice, such a distributed framework requires to address two main challenges. i) Robust optimization techniques are required to guarantee the power balance and other system constraints to be satisfied under the worst-case scenario. ii) Accurate prediction intervals should be estimated for load and renewable energy supply variations as those variations affect the optimality and the feasibility of the approach. The proposed research work include two main packages (WP1 and WP2) which attempt to address those two key features to support a reliable distributed energy trading platform.