Conference Agenda

In this paper, we study the pricing structures of virtual power purchase agreement (Virtual PPA) by Monte Carlo Simulations. Virtual PPA is one of the most effective ways for corporates to engage the carbon emission reduction strategy. A virtual PPA is a contract for difference (CfD, i.e. a derivative contract) between power buyer and renewable energy generator which settle based on the difference between an agreed strike price and a quoted electricity wholesale price. Due to the variation of the electricity price in long-term period (i.e. 10 years to 20 years for virtual PPA contract), thousands of electricity price paths should be simulated by Monte Carlo methods to understand the risk structure of net present value from the viewpoints of power buyer and renewable energy generator based on a specific long-term electricity forward curve as a base scenario. A specific algorithm using Monte Carlo method is developed to estimate the volatility of the electricity price for future 20 years. The granularity of Japan Electric Power Exchange (JEPX) spot price is 30-minute. Several VPPA pricing structures with strike price scenarios including fixed-price nominal PPA, fixed with escalation (stepped), fixed with inflation indexation. The analysis results can provide concrete quantitative evidences for strike price negotiation between power buyer and renewable energy generator.

The advent of 3D printing technology has facilitated new avenues for collaborative product design endeavors between manufacturers and consumers. The manufacturers now leverage the 3D Printing-as-a-Service (3DaaS) model by renting out 3D printers. In this paper, we study a game theoretic model setup and attempt to address the following research questions: What pricing model is suitable for offering 3DaaS? How do factors such as the degree of design customization and complexity impact the pricing strategy employed by the 3DaaS firm? Our findings indicate that when customers' influence on product quality is either high or low, the pay-per-build pricing model outperforms the fixed-fee pricing model. Additionally, we observe that in cases where customers frequently print intricate product designs, the firm may opt for the pay-per-build pricing model, contingent upon a low likelihood of design failure for these intricate structures.

We will describe our pricing system setup for an US industrial spare part vendor. The objective is to maximize the profit by controlling the daily prices for approx. 1M products. Most products are niche products, e.g. 70% of the products sold in a given months, had no sale in the months before.

Our system uses the full web traffic and soft-conversion steps, to maximize the expected profit of a user. We developed the concept of 'conversion scores' to faster evaluate price tests at scale. In this talk we will explain the core system setup and the main challenges we faced in developing this pricing system.

We suggest a dynamic pricing model for selling "empty seats" -- seat reservations for extra space in addition to regular reservations. Such extra space tickets share the resources of the main product and are viewed as a significant revenue-generating opportunity when coaches, trains, or airplanes frequently depart with many empty seats.

We formulate the problem of a transportation company that sells tickets in the same

compartment (1) without a seat reservation, (2) with a seat reservation, and (3) with a seat reservation and extra space as a Markov decision process (MDP). To address the resulting curse of dimensionality, we follow two approaches.

First, we state upper bounds based on a deterministic approximation and approximate linear programming (ALP). We show that under the Nested Logit demand model, the subproblem of the ALP row generation problem is tractable with an objective function that is convex in the action and linear in the components of the state. Corresponding decomposition approaches provide even tighter bounds.

Second, we provide policies based on these bounds. To allow for more general basis

functions, we also discuss a policy based on simulation-based approximate dynamic

programming. An extensive numerical study shows that the approaches are competitive with general approaches from literature and the best approach is the simplified decomposition. It has an excellent revenue vs. run time trade-off, rendering it a clear first choice ready for application to real-world instances. We also quantify the extra revenue potential when offering extra seats, which is particularly large in low demand settings.