TL;DR
This paper explores how different electricity market designs influence individual investment decisions in distributed solar panels. It finds that a market differentiating solar from grid energy leads to socially optimal investment, while a basic single-product market results in under-investment, and a contract-based market can lead to over-investment. These insights are crucial for designing effective policies and market structures to encourage renewable energy adoption.
What Problem Does This Paper Address?
The increasing adoption of distributed solar panels (like rooftop solar) is reshaping electricity grids. However, individual investment decisions in these systems are complex, influenced by expected future revenues from participating in short-term electricity markets. These revenues, in turn, depend on the market's equilibrium prices and allocations, which are themselves shaped by the total capacity of solar panels already installed. This creates a feedback loop: market outcomes influence investment, and aggregate investment influences market outcomes. Understanding this relationship is vital for grid operators, policymakers, and market designers to predict and guide the growth of distributed solar. The paper specifically asks: How do different short-term market mechanisms affect the long-term equilibrium of distributed solar panel investment?
Methods
The authors employ a non-atomic game framework to model the investment decisions of a large population of small, price-taking investors. This approach allows them to analyze how the aggregate investment capacity emerges from individual choices without needing to model each investor separately. They develop a unified market-modeling framework that can represent various short-term electricity market designs, accommodating features like ex-ante (contract-based) vs. ex-post (real-time) trading and product differentiation.
They then apply this framework to analyze three specific market mechanisms:
- Single-product real-time market (srt): Solar and grid electricity are treated as the same commodity.
- Product-differentiated real-time market (prt): Solar energy is distinct from grid energy, allowing consumers to value it differently.
- Contract-based market (cb): Investors commit to providing capacity ex-ante, based on expected generation.
For each mechanism, they derive analytical expressions for short-term market equilibria and then use these to characterize the long-term Nash equilibrium aggregate panel capacity.
Datasets
The paper uses solar irradiation data from San Francisco, California, from 2000-2022 to model solar generation density. Consumer willingness-to-pay data is sourced from existing literature and adjusted for inflation.
Key Insights and Ideas
- Market Design Matters Dramatically: The way electricity is traded significantly impacts how much solar capacity gets installed, and not always in a way that benefits society as a whole.
- Product Differentiation is Key for Social Optimality: Treating solar energy as a distinct product, allowing consumers to express a premium for it, leads to the "socially optimal" level of investment. This means the total benefit to society (consumers and investors) is maximized.
- Basic Markets Lead to Under-Investment: A simple market where solar and grid power are indistinguishable consistently results in less solar capacity being installed than is socially optimal. This is because the market fails to capture the extra value consumers place on clean solar energy.
- Contracts Can Lead to Over-Investment: When consumers have only a small willingness to pay a premium for solar, contract-based markets (where capacity is traded ex-ante) can incentivize more investment than is socially optimal. This might be beneficial if it drives down future costs through economies of scale.
- The Feedback Loop is Crucial: The study highlights the interconnectedness of short-term market operations and long-term investment decisions, emphasizing that understanding this loop is essential for effective policy.
Key Results
- Product-differentiated real-time market (prt): Always supports socially optimal investment.
- Single-product real-time market (srt): Consistently results in under-investment compared to the social optimum.
- Contract-based market (cb): Leads to over-investment when consumers' extra valuations for solar energy are small. This effect can persist even with larger premiums, as shown in numerical experiments.
- No Solar Premium (ε=0): When consumers don't value solar energy more than grid energy, all three market mechanisms result in the same equilibrium investment capacity.
Limitations and Future Work
- Homogeneous Investors (mostly): While the model uses a non-atomic game to represent many small investors, it often assumes homogeneous investor preferences and costs for analytical tractability. The paper does extend to heterogeneous operation periods but not fully heterogeneous investors.
- Simplified Market Rules: The analysis focuses on competitive equilibrium (CE) and Nash equilibrium (NE) and abstracts away detailed price discovery mechanisms. It also assumes uniform buy/sell prices within each market.
- Future Work: The authors suggest relaxing assumptions like uniform pricing, incorporating investor risk preferences, and exploring more complex market designs. They also note that their numerical results suggest the contract-based market's over-investment tendency persists even when solar premiums are not small, which warrants further investigation.
Summary updated