Long-term infrastructure investment planning and policy analysis for the electricity sector in Small Island Developing States: Case for Jamaica
thesisposted on 05.08.2020, 19:03 authored by Travis Renaldo AtkinsonTravis Renaldo Atkinson
Energy sector transformation is of interest to policy makers and energy researchers. Critical to this transformation is efficient (i.e. least-cost) infrastructure investment planning for new generation and transmission infrastructure investments. Similarly, energy policies designed to encourage low carbon electricity generation have fueled much of the transformation globally over the past two decades. However, knowledge gaps remain with respect to the unique economic and geographic features of Small Island Developing States (SIDS); recommendations from previous studies often have limited applicability to the SIDS context. This dissertation addresses these concerns, contributing to our understanding of least-cost planning methods for new infrastructure investments as well as energy policies appropriate for small, isolated and often heavily indebted nations. The island of Jamaica is used as a case study to gain insights more applicable to the broader SIDS context.
The first problem this dissertation addresses is the impact of simultaneously planning for generation and transmission infrastructure instead of sequentially optimizing these decisions, as is commonly done. Energy infrastructure planning in SIDS treats transmission infrastructure as an afterthought once generation investments have been determined, potentially leading to sub-optimal investments. Using a dynamic optimization model of generation and transmission infrastructure, we find that it is more cost effective to co-optimize generation and transmission investments. The substitutability between local generation and remote generation, facilitated by transmission infrastructure, underpins this result.
The second empirical problem we address is the impact of loop flow on optimal infrastructure investment decisions. The Energy Information Agency (EIA) defines loop flow as “the movement of electric power from generator to load by dividing along multiple parallel paths; it especially refers to power flow along an unintended path that loops away from the most direct geographic path or contract path” (EIA, n.d.). We find no evidence that loop flow affects optimal investment decisions in Jamaica. We attribute this to an abundance of transmission capacity and the relative simplicity of Jamaica’s network design. Results may differ for other SIDS with different starting configurations.
The third problem this dissertation addresses centers on energy policy. We quantify the cost to the Jamaican society under four different policy scenarios: a renewable portfolio standard (RPS) of 30% by year 2030, a carbon tax, a production tax credit and an investment subsidy for specific renewable energy resources (solar and wind). We find that if the decision makers’ primary concern is reducing carbon emissions, a carbon tax is the economically efficient choice (of the four options); an RPS has the second-lowest cost to society. Assessing the tradeoffs associated with each option, a carbon tax is efficient but increases the average annual cost of electricity. If, however, the decision makers’ primary objective is energy independence and not carbon emissions reduction, then the RPS may be a better alternative than a carbon tax.
Collectively, this dissertation demonstrates a method for improving long-term planning in the electricity sector in SIDS. It also quantifies the cost to society of implementing a menu of carbon mitigating policies, removing the ambiguity that persists in energy policy setting. Not only does this dissertation advance the energy economic literature by specifically addressing the economic and geographic features of SIDS, but we make our data and program files freely accessible. This is one measure that helps to overcome the data limitation hurdle that is a main contributor to the dearth of energy economics research more applicable to SIDS.